0:00:02.651,0:00:03.900
Asaf Bartov: Testing, testing.

0:00:10.036,0:00:12.640
Is this heard in the room?

0:00:15.190,0:00:15.690
Testing.

0:00:22.620,0:00:24.930
Hello, everyone.

0:00:24.930,0:00:29.460
This is a gentle[br]introduction to Wikidata

0:00:29.460,0:00:31.922
for absolute beginners.

0:00:31.922,0:00:34.130
If you're an absolute[br]beginner, if you've never heard

0:00:34.130,0:00:38.210
of Wikidata, or if you've heard[br]of Wikidata but don't quite get

0:00:38.210,0:00:41.360
it, don't know what it's[br]good for, have only used it

0:00:41.360,0:00:43.880
for inter-wiki links--

0:00:43.880,0:00:46.247
if you're anywhere[br]on this range,

0:00:46.247,0:00:47.330
you're in the right place.

0:00:50.990,0:00:52.040
My name is Asaf Bartov.

0:00:52.040,0:00:54.590
I work for the[br]Wikimedia Foundation,

0:00:54.590,0:00:59.790
and I am a Wikidata enthusiast.

0:00:59.790,0:01:05.620
So the first thing I want to[br]say is that you are lucky.

0:01:05.620,0:01:10.540
You are lucky because[br]Wikidata is already

0:01:10.540,0:01:15.415
and is quickly becoming even[br]more of an important research

0:01:15.415,0:01:21.730
tool for anyone who's[br]trying to ask questions

0:01:21.730,0:01:25.030
about large amounts[br]of information.

0:01:25.030,0:01:29.770
It will become more and more[br]used across the humanities,

0:01:29.770,0:01:33.460
in particular, because of the[br]things that it's able to do,

0:01:33.460,0:01:37.090
some of which we will[br]demonstrate shortly.

0:01:37.090,0:01:40.750
And you are lucky because you[br]get to find out about it now

0:01:40.750,0:01:43.400
before most of the world.

0:01:43.400,0:01:49.120
So by the end of this talk,[br]you will be a Wikidata hipster

0:01:49.120,0:01:51.250
because you'll be[br]able to say, oh yeah.

0:01:51.250,0:01:53.470
I knew about Wikidata[br]before it was cool.

0:01:56.090,0:02:00.370
So before we actually[br]visit Wikidata,

0:02:00.370,0:02:08.620
I want to share two key problems[br]that Wikidata seeks to solve

0:02:08.620,0:02:12.940
and which would help us[br]understand why it exists.

0:02:12.940,0:02:17.640
The first problem is that[br]have of dated data, that

0:02:17.640,0:02:20.880
is data that is out of date.

0:02:20.880,0:02:23.960
And this is apparent[br]on Wikipedia

0:02:23.960,0:02:27.870
across our free[br]knowledge encyclopedias.

0:02:27.870,0:02:32.160
Data on Wikipedia is[br]not always up to date.

0:02:32.160,0:02:37.470
And the more obscure[br]it is, the more likely

0:02:37.470,0:02:40.280
it is not to be up to date.

0:02:40.280,0:02:49.360
So the Polish Wikipedia may have[br]an article about a small town

0:02:49.360,0:02:55.480
in Argentina, and that article[br]will include information

0:02:55.480,0:03:00.910
about that town like population[br]size, name of the mayor.

0:03:00.910,0:03:04.580
And that information,[br]ideally, was

0:03:04.580,0:03:08.540
correct at the time the article[br]was created on the Polish

0:03:08.540,0:03:10.370
Wikipedia--

0:03:10.370,0:03:13.760
maybe translated[br]from another wiki.

0:03:13.760,0:03:17.900
But then how likely is[br]it to be kept up to date?

0:03:17.900,0:03:20.960
How likely is it that the[br]Polish Wikipedia would give us

0:03:20.960,0:03:25.880
the correct and latest numbers[br]or data about the population

0:03:25.880,0:03:28.370
size of that town[br]or the mayor, right?

0:03:28.370,0:03:31.720
So this is the kind of data[br]that does go out of date, right?

0:03:31.720,0:03:34.250
Every few years--[br]five, 10 years--

0:03:34.250,0:03:37.850
there is a census, and now there[br]are new population figures.

0:03:37.850,0:03:42.440
Now the census in Argentina will[br]be made available in Argentina

0:03:42.440,0:03:45.500
in Spanish, probably,[br]which brings us

0:03:45.500,0:03:48.710
to another component of the[br]problem of dated data, which

0:03:48.710,0:03:53.810
is there are no obvious[br]triggers for updating the data.

0:03:53.810,0:03:58.520
So the Polish Wikipedian[br]is not sent an email

0:03:58.520,0:04:00.680
by the Argentinean[br]government saying, hey,

0:04:00.680,0:04:01.820
we have a new census.

0:04:01.820,0:04:05.420
There are new population numbers[br]for you to update on Wikipedia.

0:04:05.420,0:04:07.550
No such email is sent.

0:04:07.550,0:04:10.146
So it's kind of[br]hard to notice when.

0:04:10.146,0:04:12.770
And of course, multiply that by[br]all the different jurisdictions

0:04:12.770,0:04:14.670
around the world.

0:04:14.670,0:04:16.610
There's no easy[br]way and notice when

0:04:16.610,0:04:17.790
your data goes out of date.

0:04:20.620,0:04:24.070
So that's difficult[br]to keep up to date.

0:04:24.070,0:04:27.940
And even if we were to receive[br]some kind of indication--

0:04:27.940,0:04:31.310
oh, there's a new[br]census in Argentina,

0:04:31.310,0:04:33.100
so a whole bunch of[br]population figures

0:04:33.100,0:04:34.960
have now gone out of date.

0:04:34.960,0:04:37.240
Updating it on the[br]Polish Wikipedia

0:04:37.240,0:04:40.090
and the French Wikipedia[br]and the Indonesian Wikipedia

0:04:40.090,0:04:44.920
and the Arabic Wikipedia is a[br]whole bunch of repetitive work

0:04:44.920,0:04:46.540
that a lot of[br]different volunteers

0:04:46.540,0:04:49.900
will need to do just for[br]that one updated piece

0:04:49.900,0:04:54.810
of information about Argentina.

0:04:54.810,0:04:57.720
So I hope this is[br]clear and resonates

0:04:57.720,0:05:01.920
with some of your experience[br]editing Wikipedia--

0:05:01.920,0:05:04.170
data that is out of[br]date or that needs

0:05:04.170,0:05:08.640
to be updated[br]manually, menially,

0:05:08.640,0:05:16.190
on a fairly frequent schedule[br]across the different countries

0:05:16.190,0:05:18.410
and data sources.

0:05:18.410,0:05:22.340
The other-- and I think[br]maybe more interesting--

0:05:22.340,0:05:26.210
shortcoming or problem[br]that I want to discuss

0:05:26.210,0:05:30.260
is what I call the[br]inflexible ways

0:05:30.260,0:05:36.020
of lateral queries, crosscutting[br]queries of knowledge.

0:05:36.020,0:05:43.980
So if I want an answer to[br]the question, what countries

0:05:43.980,0:05:48.740
in the world export rubber--

0:05:52.300,0:05:54.790
that's a reasonable[br]question, right?

0:05:54.790,0:05:57.460
That information[br]is on Wikipedia.

0:05:57.460,0:05:58.630
Do you agree?

0:05:58.630,0:06:00.640
If you go to[br]Wikipedia and read up

0:06:00.640,0:06:05.560
about Brazil, about Peru, about[br]Germany, somewhere in there--

0:06:05.560,0:06:09.010
maybe a sub-article called[br]Economics of Brazil--

0:06:09.010,0:06:13.600
you will find the main[br]exports of that country.

0:06:13.600,0:06:15.400
And you can find[br]out whether or not

0:06:15.400,0:06:16.930
that country exports rubber.

0:06:16.930,0:06:19.994
But what if I don't want[br]to go country by country

0:06:19.994,0:06:21.160
looking for the word rubber?

0:06:21.160,0:06:22.090
I just want an answer.

0:06:22.090,0:06:25.540
What are the countries[br]that export rubber?

0:06:25.540,0:06:28.360
Even though that[br]information is in Wikipedia,

0:06:28.360,0:06:29.680
it's hard to get at.

0:06:29.680,0:06:31.680
It's hard to query.

0:06:31.680,0:06:35.770
Now, you may say, well, that's[br]what we have categories for,

0:06:35.770,0:06:36.270
right?

0:06:36.270,0:06:39.820
Categories are a way to[br]cut across Wikipedia.

0:06:39.820,0:06:45.110
So if someone made a[br]category called rubber

0:06:45.110,0:06:48.380
exporting countries, then[br]you can go to that category

0:06:48.380,0:06:51.560
and see a list of countries[br]that export rubber.

0:06:51.560,0:06:53.390
And if nobody has[br]made it yet, well, you

0:06:53.390,0:06:56.990
can create that category and,[br]with a kind of one-time effort,

0:06:56.990,0:06:59.730
populate that category,[br]and you're done.

0:06:59.730,0:07:01.970
Well, yes.

0:07:01.970,0:07:04.250
That's still not[br]very convenient.

0:07:04.250,0:07:06.980
But also, it's still[br]very, very limited,

0:07:06.980,0:07:12.380
because what if I only want[br]countries that export rubber

0:07:12.380,0:07:15.950
and have a democratic[br]system of government,

0:07:15.950,0:07:18.770
or any other kind of[br]additional condition

0:07:18.770,0:07:20.510
that I would like[br]to add to this?

0:07:20.510,0:07:22.230
Or take a completely[br]different example.

0:07:22.230,0:07:26.750
What if I want to know[br]which Flemish town had

0:07:26.750,0:07:31.510
the most painters born in it?

0:07:31.510,0:07:34.480
There's a ton of[br]Flemish painters.

0:07:34.480,0:07:37.870
Most of them were[br]born somewhere.

0:07:37.870,0:07:39.685
We could theoretically,[br]just you know,

0:07:39.685,0:07:43.900
look up all the birthplaces[br]of all the Flemish painters

0:07:43.900,0:07:46.900
and tally up the[br]numbers and figure out

0:07:46.900,0:07:51.610
what is the place where the[br]most Flemish painters come from?

0:07:51.610,0:07:53.050
I don't know the answer to that.

0:07:53.050,0:07:55.420
It would be nice to be[br]able to get that answer.

0:07:55.420,0:07:57.610
Again, the data is in Wikipedia.

0:07:57.610,0:08:00.400
Those birthplaces are[br]listed in the articles

0:08:00.400,0:08:01.636
about those painters.

0:08:01.636,0:08:05.710
But there's no easy way[br]to get that information.

0:08:05.710,0:08:13.420
What if I want to ask, who are[br]some painters whose father was

0:08:13.420,0:08:14.245
also a painter?

0:08:16.840,0:08:18.500
That's a thing[br]that exists, right?

0:08:18.500,0:08:22.630
Some painters are[br]sons of painters.

0:08:22.630,0:08:26.560
You know, Bruegel comes to[br]mind as an obvious example.

0:08:26.560,0:08:28.240
But there's a bunch[br]of others, right?

0:08:28.240,0:08:29.380
So who are those people?

0:08:29.380,0:08:30.930
What if I want to[br]ask that question?

0:08:30.930,0:08:33.400
That's the kind of question[br]that not only Wikipedia

0:08:33.400,0:08:34.600
doesn't answer today.

0:08:34.600,0:08:41.500
If you walk to your friendly[br]university library reference

0:08:41.500,0:08:45.010
desk and say,[br]hello, I would like

0:08:45.010,0:08:49.290
a list of painters whose[br]father was also a painter,

0:08:49.290,0:08:52.820
how would that[br]librarian help you?

0:08:52.820,0:08:57.960
There's no easy way to get an[br]answer to a question like that.

0:08:57.960,0:09:01.100
What if you only want[br]a list of painters

0:09:01.100,0:09:05.870
who were immigrants, painters[br]who lived somewhere else

0:09:05.870,0:09:08.240
than where they were born?

0:09:08.240,0:09:09.770
There's no book.

0:09:09.770,0:09:11.720
I guess maybe there[br]is, but you know,

0:09:11.720,0:09:15.590
it's not obvious that there's a[br]ready resource that says, list

0:09:15.590,0:09:17.840
of painters who are immigrants.

0:09:17.840,0:09:19.910
And the librarian would[br]probably refer you

0:09:19.910,0:09:22.760
to a book on the shelf[br]called, I don't know,

0:09:22.760,0:09:24.200
The Complete[br]Dictionary of Flemish

0:09:24.200,0:09:26.300
Painters and go,[br]look up the index,

0:09:26.300,0:09:28.520
you know, and if you[br]see a similar surname,

0:09:28.520,0:09:29.910
maybe they're father and son.

0:09:29.910,0:09:35.000
And kind of cobble together[br]the answer on your own.

0:09:35.000,0:09:37.100
The reason I'm comparing[br]this to a library

0:09:37.100,0:09:42.170
is to show you that this is a[br]kind of question that is not

0:09:42.170,0:09:46.760
readily satisfiable today.

0:09:46.760,0:09:50.240
Now, these questions may[br]sound contrived to you.

0:09:50.240,0:09:52.460
You may say to[br]yourself, well, you

0:09:52.460,0:09:54.860
know, painters who are also[br]sons of painters, yeah.

0:09:54.860,0:09:57.680
You know, that[br]never occurred to me

0:09:57.680,0:09:59.610
as a question I[br]might care about.

0:09:59.610,0:10:01.850
But I want to invite[br]you to consider

0:10:01.850,0:10:06.380
that this kind of question,[br]questions like that question,

0:10:06.380,0:10:09.260
may well be questions[br]you do care about.

0:10:09.260,0:10:12.740
And I also want to suggest[br]that the fact it is so nearly

0:10:12.740,0:10:16.250
impossible, the fact that[br]there's no obvious way

0:10:16.250,0:10:19.250
to ask that kind[br]of question today,

0:10:19.250,0:10:21.200
is partly responsible[br]to your not

0:10:21.200,0:10:22.970
coming up with those[br]questions, right?

0:10:22.970,0:10:25.850
We tend to be limited[br]by the possible.

0:10:25.850,0:10:30.080
You know, until human[br]flight was made possible,

0:10:30.080,0:10:32.840
it did not occur to anyone[br]to say, oh yeah, by this time

0:10:32.840,0:10:34.430
next week I will[br]be in Australia,

0:10:34.430,0:10:36.630
because that was[br]just impossible.

0:10:36.630,0:10:38.587
But when flight is[br]possible, there's

0:10:38.587,0:10:40.670
all kinds of things that[br]suddenly become possible,

0:10:40.670,0:10:42.740
and there's all[br]kinds of needs that

0:10:42.740,0:10:46.430
arise based on the[br]availability of resources

0:10:46.430,0:10:48.600
to fulfill those needs.

0:10:48.600,0:10:54.120
So many of these research[br]questions, compound lateral

0:10:54.120,0:10:58.520
cross-cutting queries, are not[br]being asked because people have

0:10:58.520,0:11:00.410
internalized the fact[br]that there is no way

0:11:00.410,0:11:05.750
to get an answer[br]to questions like,

0:11:05.750,0:11:13.270
what is the most popular first[br]name among British politicians?

0:11:13.270,0:11:14.520
I just made that up, you know?

0:11:14.520,0:11:15.340
Is it John?

0:11:15.340,0:11:16.510
Maybe.

0:11:16.510,0:11:19.030
Maybe it's William,[br]for whatever reason.

0:11:19.030,0:11:22.030
You know, these are the kinds[br]of questions we don't routinely

0:11:22.030,0:11:25.855
ask because we know that it's[br]like, who are you going to ask?

0:11:25.855,0:11:28.330
How are you going to[br]get an answer to that?

0:11:28.330,0:11:36.040
So this problem of not having[br]very flexible ways of querying

0:11:36.040,0:11:38.220
the data that we already have--

0:11:38.220,0:11:41.230
in Wikipedia, in[br]Wikisource, elsewhere--

0:11:41.230,0:11:45.060
is a significant limitation.

0:11:45.060,0:11:50.880
So these two key problems[br]have one solution.

0:11:50.880,0:11:55.500
And that is an editable,[br]central storage

0:11:55.500,0:12:00.510
for structured and[br]linked data on a wiki,

0:12:00.510,0:12:05.160
under a free license, which[br]is a very long way of saying

0:12:05.160,0:12:07.290
Wikidata.

0:12:07.290,0:12:08.470
That is Wikidata.

0:12:08.470,0:12:11.190
Wikidata is an editable,[br]central storage

0:12:11.190,0:12:15.840
for structured and[br]linked data on a wiki,

0:12:15.840,0:12:17.700
under a free license.

0:12:17.700,0:12:22.590
So let's take this[br]apart and unpack it.

0:12:22.590,0:12:24.820
First of all, it's[br]a central storage.

0:12:24.820,0:12:27.660
This relates to the[br]first problem, right?

0:12:27.660,0:12:34.370
If we had one place containing[br]data like population size,

0:12:34.370,0:12:38.270
we would be able to update[br]that one place and then have

0:12:38.270,0:12:42.260
all of the different Wikipedias[br]draw the data from that one

0:12:42.260,0:12:45.320
place so that we wouldn't[br]have to manually,

0:12:45.320,0:12:49.980
repetitively update it across[br]our hundreds of projects.

0:12:49.980,0:12:53.690
So having central storage[br]makes, I hope, kind

0:12:53.690,0:12:57.230
of immediate, intuitive sense.

0:12:57.230,0:13:02.840
But what do I mean by[br]structured and linked data?

0:13:02.840,0:13:10.120
So structured data means[br]that each datum, each piece--

0:13:10.120,0:13:15.880
individual piece-- of data[br]is managed on its own,

0:13:15.880,0:13:19.660
is identified and[br]defined on its own,

0:13:19.660,0:13:21.040
as distinct from Wikipedia.

0:13:21.040,0:13:22.990
Wikipedia has articles.

0:13:22.990,0:13:27.190
The article about Brazil[br]includes a ton of data,

0:13:27.190,0:13:31.570
all kinds of information,[br]and it's presented as text,

0:13:31.570,0:13:34.270
as several paragraphs--[br]several pages--

0:13:34.270,0:13:36.540
of text, right?

0:13:36.540,0:13:41.460
Now, we do have an[br]approximation of structured data

0:13:41.460,0:13:43.580
on Wikipedia.

0:13:43.580,0:13:45.300
If you've browsed[br]Wikipedia a little,

0:13:45.300,0:13:49.100
you've noticed that we often[br]have an info box, what we

0:13:49.100,0:13:50.750
call an info box on Wikipedia.

0:13:50.750,0:13:55.220
That's the table on the right[br]side if it's a left to right

0:13:55.220,0:13:57.200
language, the table[br]on the right side

0:13:57.200,0:14:02.270
that has information that[br]is easy to tabulate, right?

0:14:02.270,0:14:08.210
So you know, birth date, birth[br]place, death date, death place,

0:14:08.210,0:14:09.710
nationality--

0:14:09.710,0:14:16.670
or if it's about a country,[br]area, population, anthem,

0:14:16.670,0:14:20.090
type of government, whatever[br]you are likely to find.

0:14:20.090,0:14:23.150
If it's a movie, then[br]you know, starring,

0:14:23.150,0:14:27.350
genre, box office receipts,[br]whatever pieces of data

0:14:27.350,0:14:29.900
are relevant to an[br]article about a movie.

0:14:29.900,0:14:34.940
So we do already kind of[br]group pieces of information

0:14:34.940,0:14:40.160
on Wikipedia into this[br]kind of structured format.

0:14:40.160,0:14:43.630
Those of you who have[br]ever looked at the source,

0:14:43.630,0:14:45.970
at what the wiki code[br]under that looks like,

0:14:45.970,0:14:49.640
know that it's only[br]semi-structured.

0:14:49.640,0:14:52.370
It looks neat and[br]organized in a table,

0:14:52.370,0:14:55.660
but really, it's just a bunch[br]of text that is put there.

0:14:55.660,0:14:57.140
It is not centralized.

0:14:57.140,0:15:00.100
Every Wikipedia has its[br]own copy of that data.

0:15:00.100,0:15:02.930
And if I go and update[br]the population size

0:15:02.930,0:15:07.070
on Spanish Wikipedia of[br]that Argentinean town,

0:15:07.070,0:15:10.190
it does not get[br]updated automagically

0:15:10.190,0:15:13.520
on the English Wikipedia or[br]the Arabic Wikipedia, right?

0:15:13.520,0:15:17.150
So the structured data that[br]we already have on Wikipedia

0:15:17.150,0:15:20.939
is not managed centrally.

0:15:20.939,0:15:22.480
The other thing[br]about structured data

0:15:22.480,0:15:29.250
is, when you have a notion of an[br]individual piece of data, that

0:15:29.250,0:15:33.390
is the cornerstone of[br]allowing the kinds of queries

0:15:33.390,0:15:34.770
that I was talking about.

0:15:34.770,0:15:40.440
That is what will allow[br]me to ask questions like,

0:15:40.440,0:15:43.470
what is the Flemish town where[br]the most painters were born,

0:15:43.470,0:15:46.650
or what are the world's[br]largest cities that

0:15:46.650,0:15:49.730
have a female mayor?

0:15:49.730,0:15:52.430
I could come up with other[br]examples all day long, right?

0:15:52.430,0:15:55.280
These are all questions[br]that you can ask,

0:15:55.280,0:15:59.390
once you break down your data[br]into individual pieces, each

0:15:59.390,0:16:02.300
of which is--

0:16:02.300,0:16:06.950
you're able to refer to each[br]of those programmatically.

0:16:06.950,0:16:10.430
The computer can[br]identify, isolate,

0:16:10.430,0:16:14.700
and calculate based on each[br]of those pieces of data.

0:16:14.700,0:16:17.060
So that's why the[br]structure is important.

0:16:17.060,0:16:22.520
Now, Wikidata is also a[br]linked data repository.

0:16:22.520,0:16:24.890
What does it mean that[br]the data is linked?

0:16:24.890,0:16:29.700
Well, it means that a single[br]piece of data can point at,

0:16:29.700,0:16:34.770
can link to another[br]whole bag of data.

0:16:34.770,0:16:43.360
So if we are describing,[br]for example, a person,

0:16:43.360,0:16:46.960
and we record the[br]single piece of data

0:16:46.960,0:16:54.820
that this person was born[br]in Salem, Massachusetts,

0:16:54.820,0:17:02.300
that single piece of data[br]links to the item about Salem,

0:17:02.300,0:17:04.060
Massachusetts[br]because, of course,

0:17:04.060,0:17:07.010
we know a lot of things[br]about that place, Salem,

0:17:07.010,0:17:07.869
Massachusetts.

0:17:07.869,0:17:09.245
So it's not just the text--

0:17:09.245,0:17:13.450
S-A-L-E-M. It's not just,[br]that's where they were born.

0:17:13.450,0:17:17.170
But it's a link to all[br]the data that we have

0:17:17.170,0:17:19.270
about Salem, Massachusetts.

0:17:19.270,0:17:24.940
If we say someone's[br]nationality is French,

0:17:24.940,0:17:26.589
that is a link to France.

0:17:26.589,0:17:30.700
That is a link to everything we[br]know about the country France.

0:17:30.700,0:17:34.150
The fact that the data[br]is linked and structured

0:17:34.150,0:17:37.630
allows not only humans,[br]but also computers

0:17:37.630,0:17:41.620
to traverse information[br]and to bring

0:17:41.620,0:17:44.950
us different pieces of[br]relevant information

0:17:44.950,0:17:49.000
programmatically, automatically,[br]based on those links.

0:17:49.000,0:17:52.000
Because it's not just[br]text, it's an actual link

0:17:52.000,0:17:56.700
to another chunk of data.

0:17:56.700,0:17:58.880
If this sounds a[br]little abstract,

0:17:58.880,0:18:01.190
it will become much[br]clearer in just a second

0:18:01.190,0:18:03.230
when we see it in action.

0:18:03.230,0:18:06.200
But the other components of[br]this little definition are,

0:18:06.200,0:18:09.650
of course, this central storage[br]of structured and linked data

0:18:09.650,0:18:12.620
needs to be editable,[br]of course, because we

0:18:12.620,0:18:14.370
need to keep it up to date.

0:18:14.370,0:18:16.460
We need to correct mistakes.

0:18:16.460,0:18:21.300
And we want it on a wiki[br]under a free license.

0:18:21.300,0:18:23.940
The free license is, of[br]course, essential to enable

0:18:23.940,0:18:30.910
reuse of that data, to enable[br]all kinds of reuse of the data.

0:18:30.910,0:18:34.060
And Wikidata, unlike[br]Wikipedia, is released

0:18:34.060,0:18:36.160
under a different free license.

0:18:36.160,0:18:41.590
Wikidata is released[br]under CC0 waiver.

0:18:41.590,0:18:44.920
That means unlike[br]Wikipedia, where

0:18:44.920,0:18:51.160
you have to attribute Wikipedia[br]when you reuse information

0:18:51.160,0:18:55.150
from Wikipedia, you do not[br]need to attribute Wikidata,

0:18:55.150,0:18:57.040
and you do not need to[br]share alike your work.

0:18:57.040,0:19:02.020
It's an unencumbered license to[br]reuse the data in any way you

0:19:02.020,0:19:03.267
want, including commercially.

0:19:03.267,0:19:05.350
You don't have to say that[br]it comes from Wikidata.

0:19:05.350,0:19:07.390
I mean, it could be nice,[br]but you don't have to.

0:19:07.390,0:19:09.280
You're under no[br]obligation to do it.

0:19:09.280,0:19:14.080
And that is important to[br]allow certain kinds of reuse

0:19:14.080,0:19:17.140
where, for example, if you're[br]building some kind of device,

0:19:17.140,0:19:20.680
you may not have a practical[br]way to give attribution.

0:19:20.680,0:19:23.920
And had we required[br]that to use Wikidata,

0:19:23.920,0:19:27.250
we would have made[br]Wikidata less reusable.

0:19:27.250,0:19:32.940
So Wikidata is unencumbered by[br]the requirement of attribution.

0:19:32.940,0:19:35.730
And of course, because[br]it's on a wiki,

0:19:35.730,0:19:40.421
we get all the benefits that we[br]are used to expect from a wiki,

0:19:40.421,0:19:40.920
right?

0:19:40.920,0:19:42.810
So it's a wiki,[br]which means, yes.

0:19:42.810,0:19:44.910
It has discussion pages.

0:19:44.910,0:19:46.500
It has revision histories.

0:19:46.500,0:19:47.620
It remembers everything.

0:19:47.620,0:19:50.610
So if you screw it up, you[br]can always go a version back.

0:19:50.610,0:19:52.380
Or if someone else[br]vandalized the content,

0:19:52.380,0:19:54.610
we can always go back,[br]just like Wikipedia.

0:19:54.610,0:19:56.880
So we get all the[br]benefits we're used to--

0:19:56.880,0:20:01.260
user talk pages, group[br]discussion pages, watch lists,

0:20:01.260,0:20:03.755
all the features that[br]we expect in a wiki.

0:20:06.740,0:20:11.170
In short, Wikidata is love.

0:20:11.170,0:20:14.100
I hope you agree with me[br]by the end of this talk.

0:20:14.100,0:20:18.580
So let's zoom in and see[br]what this structured data

0:20:18.580,0:20:21.420
looks like.

0:20:21.420,0:20:29.460
So structured data on Wikidata[br]is collected in statements.

0:20:29.460,0:20:31.930
And statements have[br]the general form

0:20:31.930,0:20:39.490
of this triple, this[br]tripartite ascription--

0:20:39.490,0:20:43.550
items, properties, and values.

0:20:43.550,0:20:46.930
Now an item is the[br]subject, is the topic

0:20:46.930,0:20:48.820
that we are trying to describe.

0:20:48.820,0:20:52.164
It can be any topic that[br]Wikipedia can cover,

0:20:52.164,0:20:53.830
and many others that[br]Wikipedia wouldn't.

0:20:53.830,0:20:57.490
So the topic, the[br]item can be Germany,

0:20:57.490,0:21:00.520
or it can be Salem,[br]Massachusetts,

0:21:00.520,0:21:03.340
or it can be the[br]concept of redemption.

0:21:03.340,0:21:04.610
It can be anything at all.

0:21:04.610,0:21:10.000
Anything you can imagine[br]describing in any way with data

0:21:10.000,0:21:11.990
can be the item.

0:21:11.990,0:21:15.430
So the item, consider[br]it like the title

0:21:15.430,0:21:17.480
of the rest of the data.

0:21:17.480,0:21:20.860
And then what do we say[br]about Salem, Massachusetts

0:21:20.860,0:21:22.330
or about Germany?

0:21:22.330,0:21:26.770
Well, that's a series of[br]properties and values,

0:21:26.770,0:21:28.450
properties and values.

0:21:28.450,0:21:32.680
The property is[br]the kind of datum,

0:21:32.680,0:21:39.770
like birth date or language[br]spoken or manner of death.

0:21:39.770,0:21:42.640
These are all real properties.

0:21:42.640,0:21:46.030
Or national anthem, if I'm[br]trying to describe a country--

0:21:46.030,0:21:47.830
these are properties.

0:21:47.830,0:21:49.880
And then they have[br]values, right?

0:21:49.880,0:21:55.740
So this person, this[br]imaginary person's place

0:21:55.740,0:21:59.640
of birth, the value of the[br]property place of birth

0:21:59.640,0:22:02.430
is Salem, Massachusetts.

0:22:02.430,0:22:06.690
So you can think about it[br]as like a government form--

0:22:06.690,0:22:09.540
or not government, just any[br]form that you're filling out--

0:22:09.540,0:22:12.420
where there are field names,[br]and then empty spaces for you

0:22:12.420,0:22:13.110
to fill out.

0:22:13.110,0:22:14.460
That's the value, OK?

0:22:14.460,0:22:18.150
So the field names[br]or the categories

0:22:18.150,0:22:19.350
are the properties, right?

0:22:19.350,0:22:22.960
So name, language,[br]occupation, date of birth--

0:22:22.960,0:22:24.420
these are all properties.

0:22:24.420,0:22:26.640
And the values are[br]the actual piece

0:22:26.640,0:22:31.391
of data, the actual[br]information that we have.

0:22:31.391,0:22:33.870
And of course,[br]different kinds of data

0:22:33.870,0:22:40.170
are relevant for describing[br]different kinds of items.

0:22:40.170,0:22:45.030
And the key in the value is it[br]can be either a literal value--

0:22:45.030,0:22:50.370
like if we're describing[br]the height of a mountain,

0:22:50.370,0:22:55.826
we might say just[br]the number 8,848.

0:22:55.826,0:22:57.325
That's the height[br]of which mountain?

0:23:01.990,0:23:04.070
Not everyone at once.

0:23:04.070,0:23:07.430
Oh, because it's meters,[br]the metric system.

0:23:07.430,0:23:08.270
Yeah, Mt.

0:23:08.270,0:23:12.390
Everest is 8,848 meters.

0:23:12.390,0:23:14.160
Yes.

0:23:14.160,0:23:15.780
Get with it, America.

0:23:15.780,0:23:17.630
The metric system.

0:23:17.630,0:23:20.930
All right, so that[br]can be a literal value

0:23:20.930,0:23:22.580
like an actual number.

0:23:22.580,0:23:28.280
Or it can be a link to an[br]item, pointing at another item.

0:23:28.280,0:23:30.890
But in this statement,[br]it is the value.

0:23:30.890,0:23:35.150
So if I'm talking about[br]Germany, the item is Germany.

0:23:35.150,0:23:39.680
And the property capital[br]city has the value Berlin.

0:23:39.680,0:23:43.130
But the value is[br]not B-E-R-L-I-N.

0:23:43.130,0:23:48.740
The value is a pointer to[br]the item Berlin, right?

0:23:48.740,0:23:51.410
That's the link.

0:23:51.410,0:23:56.671
So a single item is described[br]by a series of such statements,

0:23:56.671,0:23:57.170
right?

0:23:57.170,0:24:01.400
There's hundreds and hundreds of[br]things I can say about Germany.

0:24:01.400,0:24:04.280
There's hundreds of things[br]I can say about a person.

0:24:04.280,0:24:06.350
And these will[br]generally take the form

0:24:06.350,0:24:08.330
of a property and a value.

0:24:08.330,0:24:11.720
By the way, some properties[br]may have more than one value.

0:24:11.720,0:24:15.920
Consider the property[br]languages spoken.

0:24:15.920,0:24:18.050
People can speak more[br]than one language, right?

0:24:18.050,0:24:20.330
So if I'm from[br]describing myself,

0:24:20.330,0:24:22.400
we can say languages spoken--

0:24:22.400,0:24:26.000
English, Hebrew,[br]Latin, whatever.

0:24:26.000,0:24:27.860
So a property can have[br]more than one value.

0:24:30.970,0:24:34.010
So if the item is[br]about a country,

0:24:34.010,0:24:38.890
it would have statements about[br]properties like population,

0:24:38.890,0:24:43.180
land area, official languages,[br]borders with, anthem,

0:24:43.180,0:24:45.070
capital city.

0:24:45.070,0:24:48.580
If I'm describing a person, I[br]have a whole mostly different

0:24:48.580,0:24:51.220
set of properties that[br]are relevant, right?

0:24:51.220,0:24:54.160
Date of birth, place of birth,[br]citizenship, occupation,

0:24:54.160,0:24:56.950
father, mother,[br]religion, notable works--

0:24:56.950,0:24:59.780
now, are all of these[br]relevant for all people?

0:24:59.780,0:25:00.970
No, of course not.

0:25:00.970,0:25:02.140
It depends.

0:25:02.140,0:25:05.220
And different items[br]about different people

0:25:05.220,0:25:08.920
will either have or not[br]have these fields, right?

0:25:08.920,0:25:12.640
So we wouldn't record religion[br]for absolutely every person.

0:25:12.640,0:25:14.200
Some people manage[br]to do without.

0:25:14.200,0:25:17.710
And also, it's not relevant[br]for a lot of people, like,

0:25:17.710,0:25:20.320
what their religion[br]happens to be.

0:25:20.320,0:25:22.840
Date of birth is generally[br]relevant for most people

0:25:22.840,0:25:24.060
that we're documenting.

0:25:24.060,0:25:29.390
So some properties kind of crop[br]up more commonly than others.

0:25:29.390,0:25:33.220
A person's height, for[br]example, is not generally

0:25:33.220,0:25:35.596
considered of[br]encyclopedic value, right?

0:25:35.596,0:25:36.970
We don't, for[br]example, if we have

0:25:36.970,0:25:40.840
an article about even a[br]really well-documented person

0:25:40.840,0:25:45.610
like Winston Churchill, does[br]Wikipedia mention his height?

0:25:45.610,0:25:47.620
I don't think it does.

0:25:47.620,0:25:50.320
Even though I'm sure[br]we could probably

0:25:50.320,0:25:52.810
find a source somewhere[br]that lists his height,

0:25:52.810,0:25:55.570
it's just not a[br]very relevant piece

0:25:55.570,0:25:57.506
of information about Churchill.

0:25:57.506,0:25:59.380
With everything else[br]that's written about him

0:25:59.380,0:26:00.796
and that we know[br]about him that we

0:26:00.796,0:26:03.460
want to include in the[br]article, a person's height

0:26:03.460,0:26:08.180
is not really something of[br]great value most of the time.

0:26:08.180,0:26:14.420
But if we are describing[br]Michael Jordan, it is relevant.

0:26:14.420,0:26:15.430
I'm dating myself.

0:26:15.430,0:26:19.230
People still know[br]Michael Jordan, right?

0:26:19.230,0:26:21.600
You know, a basketball[br]player, that's

0:26:21.600,0:26:24.204
when height is very[br]relevant, right?

0:26:24.204,0:26:25.620
That's one of the[br]first things you

0:26:25.620,0:26:28.020
say when you're describing[br]a basketball player,

0:26:28.020,0:26:31.380
is list their height.

0:26:31.380,0:26:33.690
So even within the[br]class of person,

0:26:33.690,0:26:36.480
some properties may be[br]more or less relevant,

0:26:36.480,0:26:38.320
depending on the context.

0:26:38.320,0:26:40.090
So let's look at some examples.

0:26:40.090,0:26:42.870
These are examples[br]of statements.

0:26:42.870,0:26:45.400
Each line is a statement.

0:26:45.400,0:26:47.130
So here's the first one.

0:26:47.130,0:26:53.270
I want to state, about the[br]item Earth, our planet.

0:26:53.270,0:26:55.760
And what I want[br]to say about Earth

0:26:55.760,0:27:00.980
is that the property[br]highest point on Earth

0:27:00.980,0:27:03.310
has the value Mt.

0:27:03.310,0:27:04.817
Everest.

0:27:04.817,0:27:05.900
Would you agree with that?

0:27:05.900,0:27:09.580
That is the highest[br]point on Earth.

0:27:09.580,0:27:11.100
That's a statement.

0:27:11.100,0:27:14.020
It says something[br]specific, one piece

0:27:14.020,0:27:15.517
of information about Earth.

0:27:15.517,0:27:17.350
Now of course, there's[br]a lot of other things

0:27:17.350,0:27:18.820
we want to say about Earth--

0:27:18.820,0:27:21.165
circumference,[br]average temperature,

0:27:21.165,0:27:22.540
I don't know, all[br]kinds of things

0:27:22.540,0:27:26.750
we can describe the planet[br]with, density, it's a galaxy,

0:27:26.750,0:27:28.250
it belongs to, all that.

0:27:28.250,0:27:30.400
But here's one piece[br]of information,

0:27:30.400,0:27:37.370
one very specific field in[br]the detailed form about Earth.

0:27:37.370,0:27:38.990
The highest point is Mt.

0:27:38.990,0:27:39.590
Everest.

0:27:39.590,0:27:41.570
Now here's a second statement.

0:27:41.570,0:27:42.920
This time Mt.

0:27:42.920,0:27:46.690
Everest itself is the item[br]that I'm describing, right?

0:27:46.690,0:27:48.590
The topic has changed.

0:27:48.590,0:27:50.120
Now I'm saying[br]something about Mt.

0:27:50.120,0:27:52.340
Everest, and what[br]I'm saying about Mt.

0:27:52.340,0:27:56.860
Everest is elevation[br]above sea level.

0:27:56.860,0:28:01.190
Sounds the same but it[br]isn't, because the highest

0:28:01.190,0:28:04.670
point on Earth answers[br]the question where,

0:28:04.670,0:28:08.090
like on the planet, what[br]is the highest point?

0:28:08.090,0:28:08.720
It's Mt.

0:28:08.720,0:28:09.630
Everest.

0:28:09.630,0:28:12.911
But how high is that highest[br]point is a different piece

0:28:12.911,0:28:13.535
of information.

0:28:13.535,0:28:14.710
Do you agree?

0:28:14.710,0:28:16.790
It's the actual altitude.

0:28:16.790,0:28:19.600
It's not where on[br]the planet it is.

0:28:19.600,0:28:21.680
So it may sound similar,[br]but these are actually

0:28:21.680,0:28:24.030
very different pieces[br]of information.

0:28:24.030,0:28:27.800
So that highest[br]point, how high is it?

0:28:27.800,0:28:31.790
Well, it's 8,848 meters high.

0:28:31.790,0:28:36.550
Now the third statement gives[br]another piece of information

0:28:36.550,0:28:37.960
about the first item.

0:28:37.960,0:28:40.870
Same item-- I could have[br]grouped them together.

0:28:40.870,0:28:42.400
Another thing I[br]know about the Earth

0:28:42.400,0:28:46.480
is that the deepest[br]point on the planet

0:28:46.480,0:28:53.050
is the Challenger Deep, part[br]of the so-called Mariana

0:28:53.050,0:28:54.760
Trench in the ocean.

0:28:54.760,0:28:56.530
So that is the deepest point.

0:28:56.530,0:28:58.180
And how deep is it?

0:28:58.180,0:29:01.384
I again use the elevation[br]above sea level.

0:29:01.384,0:29:03.550
That's the name of the[br]property even though it's not

0:29:03.550,0:29:04.750
above sea level.

0:29:04.750,0:29:08.260
I have a negative value because[br]the elevation of the Challenger

0:29:08.260,0:29:13.700
Deep is minus 11[br]kilometers, more or less.

0:29:13.700,0:29:14.200
All right?

0:29:14.200,0:29:15.620
So these are statements.

0:29:15.620,0:29:18.820
These are four individual[br]pieces of data.

0:29:18.820,0:29:21.160
And I could also[br]look at it this way.

0:29:21.160,0:29:25.210
Maybe that's closer to the[br]government form example

0:29:25.210,0:29:26.620
that I was giving, right?

0:29:26.620,0:29:29.190
So I want to say[br]something about Earth.

0:29:29.190,0:29:30.760
What do I want to say?

0:29:30.760,0:29:33.580
Two things-- highest point.

0:29:33.580,0:29:36.760
That's the field,[br]that's the property,

0:29:36.760,0:29:37.780
and this is the value.

0:29:37.780,0:29:39.190
The highest point is Mt.

0:29:39.190,0:29:40.240
Everest.

0:29:40.240,0:29:42.880
The deepest point[br]is Challenger Deep.

0:29:42.880,0:29:46.450
And then I have things to[br]say about Challenger Deep--

0:29:46.450,0:29:49.630
the property of elevation[br]above sea level, the value

0:29:49.630,0:29:52.280
is minus 11 kilometers.

0:29:55.900,0:30:00.600
Now here's yet another[br]view of the same data

0:30:00.600,0:30:04.530
once more, with numeric IDs.

0:30:04.530,0:30:08.150
So this is the same information,[br]the same four statements.

0:30:08.150,0:30:13.020
But this time, in[br]addition to using words,

0:30:13.020,0:30:21.270
I'm also including weird[br]numbers following either Q or P.

0:30:21.270,0:30:25.890
So P stands for property.

0:30:25.890,0:30:30.330
So the highest point[br]property is P610.

0:30:30.330,0:30:34.216
And the deepest point[br]property is P1589.

0:30:34.216,0:30:35.340
What do these numbers mean?

0:30:35.340,0:30:36.985
They don't mean anything at all.

0:30:36.985,0:30:37.860
They're just numbers.

0:30:37.860,0:30:39.760
They're just sequential numbers.

0:30:39.760,0:30:42.600
And if I create a new[br]Wikidata item right now,

0:30:42.600,0:30:46.020
it'll get just the[br]next available number.

0:30:46.020,0:30:47.790
So they're just numbers.

0:30:47.790,0:30:49.080
So P stands for property.

0:30:49.080,0:30:51.480
What does Q stand for?

0:30:51.480,0:30:53.460
Does anyone know?

0:30:53.460,0:30:58.500
It's a trick question[br]because it's hard to guess.

0:30:58.500,0:31:01.896
But the principal[br]architect of Wikidata,

0:31:01.896,0:31:07.860
a Wikipedian named Danny[br][INAUDIBLE] and data scientist,

0:31:07.860,0:31:10.950
is married to a lovely[br]lady named [INAUDIBLE]

0:31:10.950,0:31:16.320
spelled with a Q. And[br]this is a loving tribute.

0:31:16.320,0:31:21.780
And she's also a Wikipedian and[br]an admin of Uzbek Wikipedia.

0:31:21.780,0:31:31.650
So Q2 is just the numeric[br]identifier of the item Earth.

0:31:31.650,0:31:36.190
And Q513 is the[br]identifier of Mt.

0:31:36.190,0:31:37.310
Everest.

0:31:37.310,0:31:42.950
You notice that we use that ID[br]across the statement, right?

0:31:42.950,0:31:48.520
So from Wikidata's[br]perspective, this

0:31:48.520,0:31:53.290
is actually what the[br]database actually contains.

0:31:53.290,0:31:55.030
What we were saying with words--

0:31:55.030,0:31:57.650
the Earth, highest[br]point, whatever--

0:31:57.650,0:31:58.540
never mind that.

0:31:58.540,0:32:03.250
Q2 has P610 with a value Q513.

0:32:03.250,0:32:06.190
That's what Wikidata[br]cares about, OK?

0:32:06.190,0:32:09.770
Now that, you'll agree,[br]is a little inaccessible.

0:32:09.770,0:32:13.120
Just these lists of numbers,[br]that's a little hard.

0:32:13.120,0:32:16.240
So Wikidata[br]understands and allows

0:32:16.240,0:32:19.690
us to continue using our words.

0:32:19.690,0:32:23.650
But actually, it gets[br]translated into numeric IDs.

0:32:23.650,0:32:25.050
Now why is this a good idea?

0:32:30.070,0:32:33.070
Why can't we just[br]say Earth or Mt.

0:32:33.070,0:32:35.120
Everest?

0:32:35.120,0:32:36.170
Any thoughts?

0:32:36.170,0:32:39.530
This is an open question.

0:32:39.530,0:32:41.540
Why is this a good[br]idea to use numbers

0:32:41.540,0:32:43.260
instead of the names of things?

0:32:47.000,0:32:51.750
Yes, because more than one[br]thing can have the same name.

0:32:51.750,0:32:52.590
What do you mean?

0:32:52.590,0:32:53.460
There's only one Mt.

0:32:53.460,0:32:54.480
Everest.

0:32:54.480,0:32:55.510
Well, yeah.

0:32:55.510,0:32:58.710
But there there's also a[br]movie called-- and probably

0:32:58.710,0:33:00.000
more than one-- called Mt.

0:33:00.000,0:33:04.080
Everest, or a TV documentary[br]literally called Mt.

0:33:04.080,0:33:06.590
Everest.

0:33:06.590,0:33:09.960
And of course, if I'm[br]describing a person named

0:33:09.960,0:33:14.930
Frank Johnson, not the only[br]Frank Johnson on the planet,

0:33:14.930,0:33:16.180
right?

0:33:16.180,0:33:17.760
But wait, you say.

0:33:17.760,0:33:20.640
On Wikipedia we deal[br]with that problem, right?

0:33:20.640,0:33:23.490
How do we deal with that[br]problem on Wikipedia?

0:33:23.490,0:33:26.270
Does anyone in[br]the audience know?

0:33:26.270,0:33:27.969
The standard way to[br]deal with the fact

0:33:27.969,0:33:30.260
that there is more than one[br]Frank Johnson in the world,

0:33:30.260,0:33:35.600
on Wikipedia, is to use[br]parentheses after the name.

0:33:35.600,0:33:39.200
So there is Frank[br]Johnson (actor)

0:33:39.200,0:33:42.620
and Frank Johnson[br](politician), for example,

0:33:42.620,0:33:44.700
if that's the distinction[br]we need to make.

0:33:44.700,0:33:48.140
So you put in parentheses[br]kind of the minimal amount

0:33:48.140,0:33:51.840
of information you need to tell[br]apart these Frank Johnsons.

0:33:51.840,0:33:54.530
What if there's two[br]politician Frank Johnsons?

0:33:54.530,0:33:58.880
Well, then you would say Frank[br]Johnson, (Delaware politician)

0:33:58.880,0:34:01.960
versus Frank Johnson[br](California politician), right?

0:34:01.960,0:34:05.210
You just put in that bit of[br]context to tell them apart.

0:34:05.210,0:34:07.640
So that's the solution[br]that Wikipedians came up

0:34:07.640,0:34:12.469
with years and years ago[br]because they did need

0:34:12.469,0:34:15.560
a unique name for the article.

0:34:15.560,0:34:18.170
You can't have two[br]articles literally called

0:34:18.170,0:34:20.790
Frank Johnson on Wikipedia.

0:34:20.790,0:34:23.570
So that's the[br]solution on Wikipedia.

0:34:23.570,0:34:28.429
But Wikidata was designed[br]much later, more than a decade

0:34:28.429,0:34:31.340
after Wikipedia, and was[br]able to kind of learn

0:34:31.340,0:34:34.520
from the experience[br]of Wikipedia, which

0:34:34.520,0:34:39.380
has tremendous experience[br]with multilingualism, much

0:34:39.380,0:34:42.870
more than most sites and[br]projects, as we know.

0:34:42.870,0:34:44.659
And so the Wikidata[br]team understood

0:34:44.659,0:34:47.840
from the get go that[br]this will be an issue,

0:34:47.840,0:34:50.989
and it's better to use[br]numbers that are unequivocally

0:34:50.989,0:34:54.800
different from each[br]other instead of labels,

0:34:54.800,0:34:57.290
instead of the actual[br]name, the actual text,

0:34:57.290,0:34:59.630
because names are not unique.

0:34:59.630,0:35:03.260
Names can change, right?

0:35:03.260,0:35:08.960
Just last year, there was a[br]big naming reform in Ukraine

0:35:08.960,0:35:13.610
and a whole bunch of towns[br]and districts were renamed.

0:35:13.610,0:35:17.330
Does that mean we should change[br]all the data that we have, like

0:35:17.330,0:35:19.550
lose all the data that we[br]have about the old name?

0:35:19.550,0:35:22.130
No, we ideally just[br]want to change the name

0:35:22.130,0:35:24.020
without breaking links.

0:35:24.020,0:35:28.550
So having the links actually[br]refer to the numbers

0:35:28.550,0:35:32.090
is one way to ensure the[br]integrity of the data,

0:35:32.090,0:35:35.360
of the links, when[br]renaming happens.

0:35:35.360,0:35:39.230
Another reason is well, even[br]if the name doesn't change,

0:35:39.230,0:35:42.230
not all humans call[br]everything the same, right?

0:35:42.230,0:35:46.180
So Earth is Earth[br]in English, but it's

0:35:46.180,0:35:48.210
[SPEAKING ARABIC] in Arabic.

0:35:48.210,0:35:49.585
It's [SPEAKING HEBREW][br]in Hebrew.

0:35:53.480,0:35:56.570
So obviously, Earth--[br]even that is not

0:35:56.570,0:36:01.920
as unambiguous or unequivocal[br]as you might think.

0:36:01.920,0:36:03.500
And so that is the[br]reason Wikidata,

0:36:03.500,0:36:07.640
which is built to be[br]multilingual from the start,

0:36:07.640,0:36:11.230
talks about numbers[br]rather than labels.

0:36:11.230,0:36:12.150
OK.

0:36:12.150,0:36:15.370
Ha, I had a whole slide[br]about that and I forgot.

0:36:15.370,0:36:17.830
Yes, so even London,[br]again, is not

0:36:17.830,0:36:20.710
just London, England, which is[br]what you were thinking about.

0:36:20.710,0:36:22.030
It's also a city in Canada.

0:36:22.030,0:36:26.260
And it's also a family[br]name, like Jack London.

0:36:26.260,0:36:27.430
It's also a movie company.

0:36:27.430,0:36:32.230
There must be some hotel[br]named London somewhere.

0:36:32.230,0:36:36.070
This is a good opportunity[br]to remind everyone

0:36:36.070,0:36:41.110
that the vast[br]majority of humankind

0:36:41.110,0:36:45.700
does not speak a[br]word of English.

0:36:45.700,0:36:48.790
That's a statistic[br]worth remembering.

0:36:48.790,0:36:55.240
The vast majority of the planet[br]does not speak English at all.

0:36:55.240,0:36:57.070
That does not[br]contradict the datum

0:36:57.070,0:37:00.070
that English is the most[br]widely spoken language.

0:37:00.070,0:37:02.860
And yet, in aggregate,[br]a majority of people

0:37:02.860,0:37:07.180
speak other languages,[br]and not English at all.

0:37:07.180,0:37:13.150
So moving swiftly on, this[br]is a pause for questions

0:37:13.150,0:37:15.610
about what I've covered so far.

0:37:15.610,0:37:17.390
Any questions in the audience?

0:37:17.390,0:37:19.450
If not, we moved to IRC.

0:37:19.450,0:37:21.042
If there are any questions--

0:37:23.880,0:37:26.891
Any questions?

0:37:26.891,0:37:27.390
No?

0:37:27.390,0:37:28.305
IRC?

0:37:28.305,0:37:29.490
Any questions?

0:37:33.580,0:37:34.180
OK.

0:37:34.180,0:37:38.170
We will have additional[br]pauses for questions later.

0:37:38.170,0:37:41.470
But enough of my hand-waving.

0:37:41.470,0:37:44.590
Let's go explore Wikidata.

0:37:44.590,0:37:49.730
So Wikidata lives[br]at wikidata.org.

0:37:49.730,0:37:59.570
And Wikidata already has[br]more than 25 million items.

0:37:59.570,0:38:05.570
That is, it collects[br]statements about more than 25

0:38:05.570,0:38:08.270
million topics.

0:38:08.270,0:38:12.170
It has many, many more[br]than 25 million statements

0:38:12.170,0:38:14.660
because many of these items[br]have dozens or hundreds

0:38:14.660,0:38:16.370
of statements.

0:38:16.370,0:38:20.720
So it documents 25[br]million things--

0:38:20.720,0:38:23.153
people, books, rivers, whatever.

0:38:26.010,0:38:28.800
Just to give us a sense[br]of how big that number is,

0:38:28.800,0:38:32.430
how many articles do we[br]have on English Wikipedia?

0:38:32.430,0:38:35.610
More than-- yes, more[br]than 5 million articles.

0:38:35.610,0:38:37.990
And that's the[br]largest Wikipedia.

0:38:37.990,0:38:41.100
So Wikidata is[br]already describing

0:38:41.100,0:38:45.450
more than five times, or[br]about five times as many items

0:38:45.450,0:38:48.460
as even our largest Wikipedia.

0:38:48.460,0:38:50.840
So obviously,[br]Wikidata contains data

0:38:50.840,0:38:56.900
about things that have no[br]article on any Wikipedia.

0:38:56.900,0:39:01.980
It is a much, much larger,[br]more comprehensive project.

0:39:01.980,0:39:04.250
All right, the second[br]thing we might notice

0:39:04.250,0:39:07.610
is, well, this looks kind[br]of like Wikipedia, right?

0:39:07.610,0:39:11.210
If we've never visited, it[br]looks kind of like Wikipedia.

0:39:11.210,0:39:13.490
It has this sidebar.

0:39:13.490,0:39:15.290
It has these buttons at the top.

0:39:15.290,0:39:17.810
It looks like it's[br]from the '90s.

0:39:17.810,0:39:18.770
Yeah.

0:39:18.770,0:39:20.900
So the reason it[br]looks like Wikipedia

0:39:20.900,0:39:24.410
is that it is a wiki running[br]on Mediawiki software.

0:39:24.410,0:39:28.430
It is running on software[br]very much like Wikipedia.

0:39:28.430,0:39:32.180
But it is running on[br]a kind of modification

0:39:32.180,0:39:34.010
of the standard wiki software.

0:39:34.010,0:39:36.170
It has an additional,[br]very important component

0:39:36.170,0:39:38.630
named Wikibase,[br]which gives it all

0:39:38.630,0:39:42.700
of its structured and[br]linked data power.

0:39:42.700,0:39:46.763
So let's start[br]exploring Wikidata.

0:39:52.830,0:39:55.770
Let's take something local--

0:39:55.770,0:39:57.530
Harvey Milk.

0:39:57.530,0:40:00.190
Harvey Milk.

0:40:00.190,0:40:03.460
What does Wikidata[br]know about Harvey Milk?

0:40:03.460,0:40:06.730
For those on YouTube[br]who may not be local,

0:40:06.730,0:40:15.580
he's a San Francisco politician[br]and gay rights activist

0:40:15.580,0:40:18.380
who was murdered in the '70s.

0:40:18.380,0:40:21.280
It was very significant in[br]the history of those struggles

0:40:21.280,0:40:22.710
in this country.

0:40:22.710,0:40:27.220
So what does Wikidata[br]tell us about Harvey Milk?

0:40:27.220,0:40:29.770
Well, the first[br]thing is it knows

0:40:29.770,0:40:34.562
that Harvey Milk is Q17141.

0:40:34.562,0:40:36.520
That's the most important[br]piece of information,

0:40:36.520,0:40:38.770
is first of all, that[br]is the identifier.

0:40:38.770,0:40:42.490
That is the item[br]number of all the data

0:40:42.490,0:40:46.150
that we will collect[br]about Harvey Milk.

0:40:46.150,0:40:50.020
The second thing you see[br]right under the title

0:40:50.020,0:40:54.730
is this line, this very,[br]very brief summary, right?

0:40:54.730,0:40:59.620
"American politician who became[br]a martyr in the gay community."

0:40:59.620,0:41:02.080
This line is the[br]description line.

0:41:02.080,0:41:04.640
So the name of the item--

0:41:04.640,0:41:05.980
this is the label.

0:41:05.980,0:41:07.450
We call it label on Wikidata.

0:41:07.450,0:41:08.740
That's the label.

0:41:08.740,0:41:10.990
And this line is[br]the description.

0:41:10.990,0:41:13.480
Now why is this[br]description important?

0:41:13.480,0:41:16.990
This is the description that[br]helps us tell this Harvey

0:41:16.990,0:41:23.230
Milk from any other Harvey[br]Milk that may exist, all right?

0:41:23.230,0:41:26.530
So again, this would[br]be useful if I'm

0:41:26.530,0:41:30.190
looking up someone with a[br]slightly more generic name.

0:41:30.190,0:41:33.910
That line will help me tell[br]apart the item about Harvey

0:41:33.910,0:41:38.860
Milk the gay activist rather[br]than Harvey Milk the film

0:41:38.860,0:41:41.750
actor, OK?

0:41:41.750,0:41:43.100
And where is it coming from?

0:41:43.100,0:41:48.690
Well, Wikidata has[br]this whole table,

0:41:48.690,0:41:52.790
as you can see, with[br]descriptions and labels

0:41:52.790,0:41:54.750
in other languages.

0:41:54.750,0:41:59.600
So Wikidata is able to refer[br]to Harvey Milk in Arabic which,

0:41:59.600,0:42:04.010
don't panic, is written[br]from right to left.

0:42:04.010,0:42:07.730
It also knows what to[br]call him in Bulgarian.

0:42:07.730,0:42:11.030
I mean, it's the same name,[br]but it's in a different script.

0:42:11.030,0:42:13.640
In French, in Hebrew,[br]and that's it?

0:42:13.640,0:42:17.960
Does it not know a name[br]for Harvey Milk in Italian?

0:42:17.960,0:42:19.760
Of course it does.

0:42:19.760,0:42:22.250
It actually has[br]labels for this person

0:42:22.250,0:42:24.435
in many, many, many languages.

0:42:24.435,0:42:30.080
It doesn't have descriptions in[br]every language, as you can see.

0:42:30.080,0:42:30.800
OK?

0:42:30.800,0:42:36.240
So why was Wikidata showing me[br]these languages and not others?

0:42:36.240,0:42:39.260
I mean, why this somewhat[br]arbitrary collection--

0:42:39.260,0:42:42.860
English, Arabic, Bulgarian,[br]German, French, and Hebrew?

0:42:42.860,0:42:45.300
Because I told it to.

0:42:45.300,0:42:50.390
So if we briefly click[br]over to my user page--

0:42:50.390,0:42:52.730
again, like every wiki,[br]you have user accounts.

0:42:52.730,0:42:53.960
You have user pages.

0:42:53.960,0:42:55.380
This is my user page.

0:42:55.380,0:42:59.750
And as you can see,[br]there's this little user

0:42:59.750,0:43:03.230
information box here called[br]a Babel box by Wikipedians,

0:43:03.230,0:43:06.610
where I list the[br]languages that I speak.

0:43:06.610,0:43:11.000
And Wikidata uses this box[br]just to kind of helpfully

0:43:11.000,0:43:12.944
show me these languages.

0:43:12.944,0:43:14.360
Of course, all the[br]other languages

0:43:14.360,0:43:19.580
are still available, as you saw,[br]by clicking the more languages.

0:43:19.580,0:43:22.940
But this is just a[br]useful little way

0:43:22.940,0:43:27.590
of getting the languages I[br]care about up there first.

0:43:27.590,0:43:29.060
By the way, this is a lie.

0:43:29.060,0:43:31.170
I don't actually[br]speak Bulgarian.

0:43:31.170,0:43:33.740
That stayed on my user page[br]because I was demonstrating

0:43:33.740,0:43:37.010
this in Bulgaria and I wanted[br]that label to show up there

0:43:37.010,0:43:38.420
during the talk--

0:43:38.420,0:43:40.250
just in case you[br]were going to tell me

0:43:40.250,0:43:43.840
a really good Bulgarian joke.

0:43:43.840,0:43:48.470
OK so for example, Hebrew[br]is my mother tongue.

0:43:48.470,0:43:51.730
And we have a Hebrew[br]label for Harvey Milk.

0:43:51.730,0:43:53.810
But we don't have a description.

0:43:53.810,0:44:00.950
So let's fix that right now by[br]clicking the edit button right

0:44:00.950,0:44:01.960
here.

0:44:01.960,0:44:05.930
I click edit, and this[br]table became editable.

0:44:05.930,0:44:09.661
And now I can very briefly[br]type a description.

0:44:22.899,0:44:24.440
AUDIENCE: Online in[br]about 20 seconds.

0:44:24.440,0:44:25.400
But can we hold it?

0:44:25.400,0:44:26.066
ASAF BARTOV: OK.

0:44:28.454,0:44:30.430
That was good timing[br]for the screen to crash.

0:44:53.642,0:44:54.142
OK?

0:44:59.082,0:45:01.800
Are we back?

0:45:01.800,0:45:02.850
OK.

0:45:02.850,0:45:03.690
Sorry about that.

0:45:03.690,0:45:07.500
So this was all about what to[br]call him in different languages

0:45:07.500,0:45:09.930
and scripts and how to[br]tell this person apart

0:45:09.930,0:45:13.590
from other people with[br]potentially the same name.

0:45:13.590,0:45:17.930
Let's scroll down and see[br]what else does Wikidata

0:45:17.930,0:45:19.680
know about this person?

0:45:19.680,0:45:24.060
So as you can see, this is[br]a list of statements, right?

0:45:24.060,0:45:25.500
This is a list of statements.

0:45:25.500,0:45:27.900
And the properties[br]are on the left,

0:45:27.900,0:45:30.340
the values are on the right.

0:45:30.340,0:45:33.870
So the first thing Wikidata[br]knows about Harvey Milk

0:45:33.870,0:45:38.520
is a very important[br]property called instance of.

0:45:38.520,0:45:39.910
Instance of.

0:45:39.910,0:45:44.690
And the property instance of[br]answers the very basic question

0:45:44.690,0:45:49.460
what kind of thing is[br]this that I'm describing?

0:45:49.460,0:45:50.870
Is it a book?

0:45:50.870,0:45:51.980
Is it a poem?

0:45:51.980,0:45:53.570
Is it a mountain?

0:45:53.570,0:45:55.520
Is it a theological concept?

0:45:55.520,0:45:57.800
No, it's a human.

0:45:57.800,0:46:00.020
It's a person, OK?

0:46:00.020,0:46:01.880
The item about Mt.

0:46:01.880,0:46:07.070
Everest will say[br]instance of mountain, OK?

0:46:07.070,0:46:10.790
This is a very[br]important property.

0:46:10.790,0:46:12.500
Why is it important?

0:46:12.500,0:46:14.630
Wouldn't anyone looking[br]at this know that this is

0:46:14.630,0:46:15.550
a human being?

0:46:15.550,0:46:16.310
Yes.

0:46:16.310,0:46:18.720
Anyone looking at[br]this will know.

0:46:18.720,0:46:23.780
But if I want a computer to[br]be able to pull information

0:46:23.780,0:46:28.160
about people, I want to[br]be able to easily exclude

0:46:28.160,0:46:30.680
all the mountains and[br]poems and other things that

0:46:30.680,0:46:33.440
are not people from my query.

0:46:33.440,0:46:37.400
So this single datum,[br]this single piece of data,

0:46:37.400,0:46:41.720
is what tells computers and[br]algorithms very clearly,

0:46:41.720,0:46:42.890
this is a human.

0:46:42.890,0:46:47.340
Things that aren't instance[br]of human are other things.

0:46:47.340,0:46:48.230
OK?

0:46:48.230,0:46:50.145
So it may sound very[br]trivial, but it's not.

0:46:50.145,0:46:51.770
It's very important[br]to have an instance

0:46:51.770,0:46:54.077
of field for Wikidata items.

0:46:54.077,0:46:55.410
All right, what else do we know?

0:46:55.410,0:46:59.360
Well, Wikidata knows about[br]an image for Harvey Milk.

0:46:59.360,0:47:02.982
Again, we can find a ton of[br]images-- or maybe not a ton,

0:47:02.982,0:47:04.940
but we can find dozens[br]of images of Harvey Milk

0:47:04.940,0:47:10.430
on Commons, on our Wikimedia[br]multimedia repository.

0:47:10.430,0:47:13.430
So why should we have a[br]single image here on Wikidata?

0:47:13.430,0:47:16.280
Again, this is[br]mostly for reusers.

0:47:16.280,0:47:18.920
If I'm building some kind of[br]tool that pulls information

0:47:18.920,0:47:21.680
from Wikidata, it's[br]nice if there's

0:47:21.680,0:47:24.680
at least one representative[br]image to kind of use

0:47:24.680,0:47:30.300
as the default or immediate[br]image for Harvey Milk

0:47:30.300,0:47:33.120
in some other reused context.

0:47:33.120,0:47:34.770
All right, sex or gender--

0:47:34.770,0:47:35.670
male.

0:47:35.670,0:47:38.790
Country of citizenship--[br]United States of America.

0:47:38.790,0:47:39.910
Given name is Harvey.

0:47:39.910,0:47:41.580
The date of birth is so and so.

0:47:41.580,0:47:44.340
The place of birth is Woodmere.

0:47:44.340,0:47:45.870
The place of death[br]is San Francisco.

0:47:45.870,0:47:48.640
The manner of death is homicide.

0:47:48.640,0:47:50.930
Wikidata knows that.

0:47:50.930,0:47:55.700
Now again, every[br]little datum like that

0:47:55.700,0:48:02.210
is the basis for later querying[br]and answering questions.

0:48:02.210,0:48:07.390
So the fact that we record the[br]manner of death of people--

0:48:07.390,0:48:09.230
or at least of some people--

0:48:09.230,0:48:11.900
will allow us later[br]to go, you know,

0:48:11.900,0:48:17.120
who are some people from[br]Belgium who died by homicide?

0:48:17.120,0:48:24.650
That's a question Wikidata can[br]answer, thanks to this field.

0:48:24.650,0:48:27.680
The other thing I mentioned[br]is that things are links.

0:48:27.680,0:48:29.680
So the place of[br]birth is Woodmere.

0:48:29.680,0:48:31.900
I don't know where[br]Woodmere is, but I

0:48:31.900,0:48:34.390
can click that and find out.

0:48:34.390,0:48:38.270
Here is the Wikidata item[br]about Woodmere, right?

0:48:38.270,0:48:41.230
It was the value in the[br]statement about Harvey Milk,

0:48:41.230,0:48:43.900
but now I'm looking at[br]the item about Woodmere.

0:48:43.900,0:48:48.047
And it turns out it's in[br]Nassau County, New York, right?

0:48:48.047,0:48:50.380
And of course, Wikidata has[br]a whole bunch of information

0:48:50.380,0:48:55.450
for me about Woodmere--

0:48:55.450,0:48:59.720
what country it's in and the[br]coordinates and the population

0:48:59.720,0:49:06.230
and the area, all the things you[br]would expect about a place, OK?

0:49:06.230,0:49:07.512
Let's get back to Harvey Milk.

0:49:10.370,0:49:13.260
So the manner of death,[br]the cause of death--

0:49:13.260,0:49:16.880
now here, Wikidata gives[br]us excellent information.

0:49:16.880,0:49:20.390
The actual cause of death[br]is ballistic trauma.

0:49:20.390,0:49:22.160
That's a professional term.

0:49:22.160,0:49:27.560
And this statement[br]has qualifiers.

0:49:27.560,0:49:30.650
So until now, I was talking[br]about triples, right?

0:49:30.650,0:49:33.260
The item has a property[br]with a certain value.

0:49:33.260,0:49:35.270
Actually, each[br]statement can also

0:49:35.270,0:49:38.030
have a number of[br]qualifiers which

0:49:38.030,0:49:45.424
add aspects of information,[br]still about that one question

0:49:45.424,0:49:46.590
that we're answering, right?

0:49:46.590,0:49:49.904
So if this property[br]answers cause of death,

0:49:49.904,0:49:51.320
it's not discussing[br]anything else.

0:49:51.320,0:49:52.880
It's not discussing languages.

0:49:52.880,0:49:54.920
It's not discussing[br]date of birth, right?

0:49:54.920,0:49:56.930
It's talking about[br]the cause of death.

0:49:56.930,0:49:59.300
But we're not just[br]saying ballistic trauma.

0:49:59.300,0:50:04.550
We're saying ballistic trauma[br]with the quantity attribute

0:50:04.550,0:50:05.660
being five.

0:50:05.660,0:50:07.550
What does that mean?

0:50:07.550,0:50:08.870
Five bullets, right?

0:50:08.870,0:50:12.780
There are five[br]ballistic traumas.

0:50:12.780,0:50:15.300
He was he was shot five times.

0:50:15.300,0:50:18.210
And he was shot by this[br]person named Dan White.

0:50:18.210,0:50:25.020
And this ballistic trauma,[br]like this actual shooting,

0:50:25.020,0:50:28.420
is itself the subject[br]of this other thing.

0:50:28.420,0:50:31.440
This is a link to a[br]whole other Wikidata

0:50:31.440,0:50:35.510
item about the Moscone-Milk[br]assassinations.

0:50:35.510,0:50:38.610
Moscone was the San[br]Francisco mayor at the time.

0:50:43.540,0:50:47.510
We'll see slightly better or[br]easier to understand examples

0:50:47.510,0:50:49.460
of qualifiers in a bit.

0:50:49.460,0:50:54.440
So if this was[br]confusing, hang on.

0:50:54.440,0:50:55.970
So he was killed by Dan White.

0:50:55.970,0:50:57.800
He spoke English.

0:50:57.800,0:50:59.960
His occupation--[br]here's an example

0:50:59.960,0:51:03.140
of a property with more[br]than one value, right?

0:51:03.140,0:51:06.260
So Milk was a politician.

0:51:06.260,0:51:09.710
But he was also a Navy[br]officer, at least for a while.

0:51:09.710,0:51:12.980
That was another thing that[br]he did during his life.

0:51:12.980,0:51:15.350
And he was a human[br]rights activist, right?

0:51:15.350,0:51:20.600
So some people are[br]writers and translators.

0:51:20.600,0:51:22.610
So people can have more[br]than one occupation.

0:51:22.610,0:51:26.310
People can speak more[br]than one language.

0:51:26.310,0:51:29.130
Here's a better[br]example of a qualifier.

0:51:29.130,0:51:35.090
So the property award received[br]has the value Presidential

0:51:35.090,0:51:37.560
Medal of Freedom.

0:51:37.560,0:51:42.570
And that award has an[br]attribute called point in time,

0:51:42.570,0:51:44.070
like when was this?

0:51:44.070,0:51:46.580
This was in 2009.

0:51:46.580,0:51:50.510
Do you see that[br]this piece of data--

0:51:50.510,0:52:04.780
2009-- is a sub-statement[br]or is subjugated

0:52:04.780,0:52:09.621
to the context of this award,[br]was the Presidential Medal

0:52:09.621,0:52:10.120
of Freedom?

0:52:10.120,0:52:13.430
It can't just kind of[br]free float in the article.

0:52:13.430,0:52:17.650
It's not that 2009 is itself[br]a meaningful thing, right?

0:52:17.650,0:52:21.550
This medal was awarded in 2009.

0:52:21.550,0:52:22.170
If

0:52:22.170,0:52:24.070
Wikidata doesn't[br]tell us, for example,

0:52:24.070,0:52:27.130
when he was a Navy officer, OK?

0:52:27.130,0:52:30.100
But if we were, for example,[br]to look that up right now

0:52:30.100,0:52:33.820
and find out that Milk was[br]a Navy officer between 1962

0:52:33.820,0:52:39.542
and 1964, we could go back[br]here to the Navy officer bit

0:52:39.542,0:52:41.010
and click edit.

0:52:41.010,0:52:44.190
This is how I edit this[br]particular little piece

0:52:44.190,0:52:45.360
of information.

0:52:45.360,0:52:49.350
And add a qualifier like this.

0:52:49.350,0:52:51.300
I click Add Qualifier.

0:52:51.300,0:52:57.660
And I could pick start[br]time and end time, right?

0:52:57.660,0:53:04.990
And then I could[br]type 1962 to 1964,

0:53:04.990,0:53:08.000
and that would be[br]teaching Wikidata.

0:53:08.000,0:53:10.660
Oh, I'm sorry, I meant to[br]do that for Navy officer.

0:53:10.660,0:53:11.230
OK.

0:53:11.230,0:53:14.800
But, you know,[br]that is the exact--

0:53:14.800,0:53:18.400
the accurate time span[br]of that statement.

0:53:18.400,0:53:22.850
So it's true to say about a[br]person, he was a Navy officer,

0:53:22.850,0:53:25.990
even if of course he wasn't a[br]Navy officer his entire life.

0:53:25.990,0:53:28.120
But it's better and[br]it's more accurate,

0:53:28.120,0:53:32.260
to say he was a Navy officer[br]between 1962 and 1964.

0:53:32.260,0:53:35.380
Don't worry, I'm[br]not saving this.

0:53:35.380,0:53:39.150
No vandalizing of[br]Wikidata in this session.

0:53:39.150,0:53:40.450
OK.

0:53:40.450,0:53:41.140
Moving on.

0:53:41.140,0:53:42.430
What else does Wikidata know?

0:53:42.430,0:53:43.960
He was educated at[br]this university.

0:53:43.960,0:53:46.970
He was a member of[br]this political party.

0:53:46.970,0:53:47.470
Right?

0:53:47.470,0:53:49.428
That's of course if[br]they're a relevant property

0:53:49.428,0:53:52.270
for a politician.

0:53:52.270,0:53:56.500
Religion, military branch,[br]what is the category on commons

0:53:56.500,0:53:58.720
that discusses this[br]item, is something

0:53:58.720,0:54:00.790
that Wikidata can tell us.

0:54:00.790,0:54:02.200
And that's it.

0:54:02.200,0:54:04.570
Now, is that everything[br]that we could possibly

0:54:04.570,0:54:07.780
say in a structured[br]way about Harvey Milk?

0:54:07.780,0:54:08.680
No.

0:54:08.680,0:54:13.570
We could probably find at[br]least a few more things to say.

0:54:13.570,0:54:17.170
We will see how to contribute[br]new information to Wikidata

0:54:17.170,0:54:19.990
in just a minute with[br]a different example.

0:54:19.990,0:54:23.360
But this-- all this was[br]a set of statements.

0:54:23.360,0:54:23.860
Right?

0:54:23.860,0:54:25.927
This was the title[br]statements here.

0:54:28.840,0:54:31.160
But at the bottom of the[br]list of statements is

0:54:31.160,0:54:34.300
another section[br]called identifiers.

0:54:34.300,0:54:36.960
And I want to spend a minute[br]talking about what that is.

0:54:36.960,0:54:43.630
So identifiers is a[br]collection of keys.

0:54:43.630,0:54:47.980
A collection of[br]IDs, or codes, that

0:54:47.980,0:54:52.890
are keys to other[br]information sources.

0:54:52.890,0:54:58.560
And a lot of Wikidata items[br]have a whole series of keys

0:54:58.560,0:55:03.030
to other databases, other[br]sites, other repositories,

0:55:03.030,0:55:08.340
that help you or a computer[br]be able to access not just

0:55:08.340,0:55:12.240
some database and look for[br]information about Harvey Milk,

0:55:12.240,0:55:16.950
but access the exact record[br]relevant to Harvey Milk.

0:55:16.950,0:55:20.280
And again, if you imagine[br]someone named John Smith,

0:55:20.280,0:55:21.690
that is really valuable, right?

0:55:21.690,0:55:23.250
If you're not just[br]told, oh yeah,

0:55:23.250,0:55:24.875
you can look at the[br]Library of Congress

0:55:24.875,0:55:27.840
for John Smith,[br]good luck with that.

0:55:27.840,0:55:30.240
Or if I tell you, go to[br]the Library of Congress

0:55:30.240,0:55:35.810
to this record for this John[br]Smith, you see the difference.

0:55:35.810,0:55:42.080
So Wikidata tells us that on[br]VIAF, which is the Virtual

0:55:42.080,0:55:44.570
International Authority File.

0:55:44.570,0:55:50.140
It's an aggregated master[br]index built by bibliographers,

0:55:50.140,0:55:52.831
by librarians, of people.

0:55:52.831,0:55:53.330
Right?

0:55:53.330,0:55:56.720
It tries to kind of aggregate[br]information about people

0:55:56.720,0:55:59.270
across library[br]catalogs everywhere.

0:55:59.270,0:56:05.120
So the VIAF ID for Harvey[br]Milk is this number.

0:56:05.120,0:56:07.340
And conveniently,[br]if I click that,

0:56:07.340,0:56:10.160
I'm not taking to[br]some Wikidata item.

0:56:10.160,0:56:13.010
I'm actually taken[br]to the relevant site.

0:56:13.010,0:56:16.760
So this took me right[br]to viaf.org, the Virtual

0:56:16.760,0:56:21.770
International Authority File,[br]directly to their record

0:56:21.770,0:56:23.310
about Harvey Milk.

0:56:23.310,0:56:23.810
All right?

0:56:23.810,0:56:27.290
And that itself leads[br]me to national catalogs

0:56:27.290,0:56:29.630
of national libraries[br]all over the world.

0:56:29.630,0:56:32.360
We won't get into the[br]things you can do with VIAF.

0:56:32.360,0:56:37.220
The point is Wikidata[br]contained the piece of thread

0:56:37.220,0:56:40.820
that I could tug on[br]to arrive directly

0:56:40.820,0:56:44.840
to that information[br]in other databases.

0:56:44.840,0:56:45.680
Yes.

0:56:45.680,0:56:49.670
And it has that for many,[br]many kinds of databases.

0:56:49.670,0:56:53.150
The BNF, for example, that's[br]the National Library of France.

0:56:53.150,0:56:56.270
And that will take me[br]to that index card.

0:56:56.270,0:56:57.320
IMDB.

0:56:57.320,0:56:58.620
We all know IMDB, right?

0:56:58.620,0:57:03.320
So here I have the key[br]to Harvey Milk in IMDB.

0:57:03.320,0:57:05.810
And this is what IMDB says[br]about Harvey Milk, right?

0:57:05.810,0:57:08.480
They have their own piece[br]of information about him,

0:57:08.480,0:57:11.590
of course, with filmography[br]and everything else.

0:57:11.590,0:57:15.140
And see, I did not have[br]to search IMDB for it.

0:57:15.140,0:57:19.070
I just had the key right[br]there waiting for me.

0:57:19.070,0:57:21.080
Now, again, this is[br]very convenient for me

0:57:21.080,0:57:24.590
as I just showed you the[br]human use case for this.

0:57:24.590,0:57:27.530
But it's even more[br]powerful in aggregate

0:57:27.530,0:57:35.450
when we allow computers to[br]traverse this network of links

0:57:35.450,0:57:36.110
between--

0:57:36.110,0:57:41.690
not just within wiki data, but[br]between data storage facilities

0:57:41.690,0:57:43.850
and repositories.

0:57:43.850,0:57:49.790
This is sometimes referred to[br]as the linked data open cloud.

0:57:49.790,0:57:52.670
Cloud, because it's multiple[br]different repositories

0:57:52.670,0:57:54.740
that are interlinked.

0:57:54.740,0:58:02.210
And Wikidata is already, and[br]to a growing extent, the Nexus,

0:58:02.210,0:58:04.460
the connection[br]point between a lot

0:58:04.460,0:58:06.780
of these different databases.

0:58:06.780,0:58:09.230
So IMDB, for example,[br]it's a good example

0:58:09.230,0:58:11.300
because it's site[br]almost everyone knows,

0:58:11.300,0:58:14.000
IMDB has information[br]about Harvey Milk.

0:58:14.000,0:58:16.670
But that information[br]does not include a link

0:58:16.670,0:58:19.140
to the French National Library.

0:58:19.140,0:58:19.645
Right?

0:58:19.645,0:58:20.770
Do you see what I'm saying?

0:58:20.770,0:58:25.550
So IMDB is a data repository[br]with IDs and allows linking.

0:58:25.550,0:58:28.100
But it does not give you[br]what Wikidata gives you which

0:58:28.100,0:58:32.850
is this kind of collection of--

0:58:32.850,0:58:36.330
it's like a junction of all[br]these different data sources.

0:58:36.330,0:58:37.910
So Wikidata is the[br]place where you

0:58:37.910,0:58:40.730
can document these[br]interrelationships

0:58:40.730,0:58:41.640
or equivalencies.

0:58:41.640,0:58:42.140
Right?

0:58:42.140,0:58:48.770
So ID, you know, 587548 on IMDB[br]is discussing the same topic

0:58:48.770,0:58:52.260
as French National[br]Library ID whatever.

0:58:52.260,0:58:55.210
Wikidata contains that[br]piece of information.

0:58:55.210,0:58:59.090
that this ID in this database[br]is about the same person

0:58:59.090,0:59:04.050
as that ID in that database.

0:59:04.050,0:59:05.290
OK.

0:59:05.290,0:59:07.420
So that's what[br]identifiers are about.

0:59:07.420,0:59:11.320
Still scrolling down the[br]Wikidata item about Harvey

0:59:11.320,0:59:15.500
Milk, we have the site links.

0:59:15.500,0:59:20.840
The site links are links[br]to Wikimedia projects

0:59:20.840,0:59:22.770
that are related to this item.

0:59:22.770,0:59:25.250
So of course there[br]are Wikipedia articles

0:59:25.250,0:59:28.880
about Harvey Milk in many,[br]many different wikipedias.

0:59:28.880,0:59:31.700
Quite a few language versions.

0:59:31.700,0:59:34.960
And there are[br]pages on Wikiquote,

0:59:34.960,0:59:36.680
one of the sister projects.

0:59:36.680,0:59:38.630
There are pages on[br]Wikiquote with some quotes

0:59:38.630,0:59:40.130
from Harvey Milk.

0:59:40.130,0:59:45.060
And there is even a page for[br]Harvey Milk on Wikisource.

0:59:45.060,0:59:45.560
Right?

0:59:45.560,0:59:47.840
So this is a collection[br]of those links.

0:59:47.840,0:59:52.760
And those of you who have maybe[br]only dealt with Wikidata data

0:59:52.760,0:59:57.290
for inter-wiki links, which[br]we used to do in the old days

0:59:57.290,0:59:59.600
manually within[br]the article text,

0:59:59.600,1:00:01.716
now we do it through[br]Wikidata, so maybe that's

1:00:01.716,1:00:03.590
the only thing you didn't[br]know about Wikidata

1:00:03.590,1:00:10.130
is how to update these[br]inter-wiki tables on Wikidata.

1:00:10.130,1:00:11.430
All right.

1:00:11.430,1:00:14.090
So that concludes[br]our little tour

1:00:14.090,1:00:18.560
of the anatomy of[br]a Wikidata page.

1:00:18.560,1:00:22.370
I will just remind you that[br]it's a wiki page, which

1:00:22.370,1:00:26.120
means it has a discussion[br]page, a talk page.

1:00:26.120,1:00:27.960
This one happens to be empty.

1:00:27.960,1:00:30.092
But, you know, if we have[br]concerns or arguments

1:00:30.092,1:00:31.550
about some of the[br]data here that is

1:00:31.550,1:00:33.290
what we would use[br]to discuss this

1:00:33.290,1:00:36.830
and to arrive at consensus.

1:00:36.830,1:00:41.760
It also has a history view just[br]like every Wikipedia article.

1:00:41.760,1:00:47.402
So you can see here[br]a list of edits.

1:00:47.402,1:00:48.860
Maybe some of you[br]have never looked

1:00:48.860,1:00:51.710
at a history page on Wikipedia,[br]so this looks overwhelming.

1:00:51.710,1:00:55.040
But every line here,[br]every entry here,

1:00:55.040,1:00:58.240
is a single edit, a single[br]revision, a single change

1:00:58.240,1:01:00.440
to this Wikidata item.

1:01:00.440,1:01:01.670
Just Harvey Milk.

1:01:01.670,1:01:04.250
And you can see at the very[br]top this edit that I just

1:01:04.250,1:01:06.680
made-- this is my[br]volunteer account

1:01:06.680,1:01:09.650
and I just made this edit,[br]and in parentheses you

1:01:09.650,1:01:10.790
can see what I did.

1:01:10.790,1:01:14.640
I added an HE,[br]Hebrew, description.

1:01:14.640,1:01:16.930
And this is the text[br]that I added in Hebrew.

1:01:16.930,1:01:17.430
Right?

1:01:17.430,1:01:21.470
So we can see who added[br]what to the Wikidata item,

1:01:21.470,1:01:24.960
just like we can do[br]the same on Wikipedia.

1:01:24.960,1:01:26.390
So we have the revision history.

1:01:26.390,1:01:27.560
We can undo edits.

1:01:27.560,1:01:30.320
We can revert, just[br]like on Wikipedia.

1:01:34.420,1:01:36.940
And what else did I[br]want to show here?

1:01:36.940,1:01:40.930
We can add an item to my[br]watch list using the star,

1:01:40.930,1:01:42.020
just like on Wikipedia.

1:01:42.020,1:01:46.670
So we have all these[br]standard wiki features

1:01:46.670,1:01:47.878
that we would come to expect.

1:01:50.440,1:01:54.270
Let's pause for questions.

1:01:54.270,1:01:58.412
Any questions about what[br]we've covered so far?

1:02:02.573,1:02:03.073
Yes.

1:02:06.950,1:02:11.345
Are attributes of statements[br]precept for the specific value?

1:02:16.640,1:02:19.830
No they're not reset.

1:02:19.830,1:02:29.760
And generally Wikidata data does[br]not enforce by default logic.

1:02:29.760,1:02:32.130
So, I mean, there's[br]nothing to prevent you

1:02:32.130,1:02:38.700
from editing the[br]item about Brazil,

1:02:38.700,1:02:42.990
and adding the property height.

1:02:46.690,1:02:50.430
Now height is not a relevant[br]property for a country.

1:02:50.430,1:02:50.970
Right?

1:02:50.970,1:02:53.880
I mean, maybe average[br]elevation, maybe.

1:02:53.880,1:02:56.400
But not just height,[br]which is used for humans

1:02:56.400,1:02:59.040
or for physical things.

1:02:59.040,1:03:02.400
So you could add that[br]property to Brazil and save it

1:03:02.400,1:03:04.650
and the wiki would not complain.

1:03:04.650,1:03:07.590
Now in the background[br]there are kind

1:03:07.590,1:03:13.020
of extra wiki outside the[br]wiki prostheses for constraint

1:03:13.020,1:03:13.710
validation.

1:03:13.710,1:03:16.050
So there are bots and[br]other processes that

1:03:16.050,1:03:17.940
run, and occasionally,[br]for example,

1:03:17.940,1:03:26.570
identify non-living things[br]with a date of birth field.

1:03:26.570,1:03:27.720
That's nonsensical.

1:03:27.720,1:03:29.010
That should not exist.

1:03:29.010,1:03:31.710
If someone mistakenly added[br]that there are processes

1:03:31.710,1:03:34.350
that would flag[br]that to be fixed.

1:03:34.350,1:03:36.690
But the wiki itself,[br]Wikidata, will not

1:03:36.690,1:03:38.550
prevent you from adding that.

1:03:38.550,1:03:41.940
And that is by design[br]to keep things flexible.

1:03:41.940,1:03:43.930
So that people don't[br]run into, oh wait,

1:03:43.930,1:03:46.560
but I can't add this[br]because nobody thought

1:03:46.560,1:03:49.830
that I would need this, maybe.

1:03:49.830,1:03:54.530
I hope that answers[br]your question.

1:03:54.530,1:03:57.290
You say helpful[br]answer, question mark.

1:03:57.290,1:03:59.510
So was it a helpful answer, or?

1:04:03.940,1:04:04.440
OK.

1:04:04.440,1:04:05.426
Yes, Eleanor.

1:04:05.426,1:04:10.707
AUDIENCE: [INAUDIBLE]

1:04:10.707,1:04:12.040
ASAF BARTOV: Excellent question.

1:04:12.040,1:04:13.030
I'll repeat it.

1:04:13.030,1:04:16.180
You ask how do I find[br]the wiki data item

1:04:16.180,1:04:18.370
number from Wikipedia.

1:04:18.370,1:04:21.580
If I'm reading about Harvey Milk[br]and I want to look at the data

1:04:21.580,1:04:23.600
how do I do that?

1:04:23.600,1:04:27.400
That is an excellent question[br]and let's skip to Wikipedia.

1:04:27.400,1:04:32.030
Conveniently I have the[br]link right here on English.

1:04:32.030,1:04:35.600
So this is the Wikipedia[br]article about Harvey Milk

1:04:35.600,1:04:42.740
and every item on Wikipedia[br]should have a wiki data

1:04:42.740,1:04:47.660
item associated with it, but it[br]doesn't happen automatically.

1:04:47.660,1:04:51.470
So if I just created[br]a page on Wikipedia

1:04:51.470,1:04:55.010
I also need to create a[br]Wikidata entity for it

1:04:55.010,1:04:57.170
if it doesn't already exist.

1:04:57.170,1:04:59.420
It could already exist[br]because it was already

1:04:59.420,1:05:01.970
covered in a different[br]language, for example.

1:05:01.970,1:05:05.390
So that was parenthetical.

1:05:05.390,1:05:09.020
But every article on Wikipedia[br]should have, here on the side,

1:05:09.020,1:05:14.270
on the side are under Tools,[br]a link called Wikidata item.

1:05:14.270,1:05:15.450
Right here.

1:05:15.450,1:05:16.160
OK.

1:05:16.160,1:05:18.110
That Wikidata data[br]item is a link

1:05:18.110,1:05:21.710
that takes you to[br]Wikidata, to the entity,

1:05:21.710,1:05:23.510
and there you find the number.

1:05:23.510,1:05:25.370
You can-- you don't[br]even have to click it.

1:05:25.370,1:05:27.830
I mean, the URL itself[br]tells you the number.

1:05:27.830,1:05:34.620
The number, you see, it's[br]wikidata.org/wiki/q17141.

1:05:34.620,1:05:35.444
OK.

1:05:35.444,1:05:36.860
So that was an[br]excellent question.

1:05:36.860,1:05:37.686
Other questions?

1:05:37.686,1:05:38.185
Yes.

1:05:41.470,1:05:44.430
Yeah, about the additional[br]attributes, the qualifiers.

1:05:44.430,1:05:46.920
So, yes, I answered[br]more generically.

1:05:46.920,1:05:49.370
But just like the[br]properties themselves

1:05:49.370,1:05:53.390
are not limited per item,[br]the qualifiers per statement

1:05:53.390,1:05:57.750
are also not[br]entirely preordained.

1:05:57.750,1:05:59.570
But there is some[br]structure to it.

1:05:59.570,1:06:03.140
I don't want to go into it[br]at great length right now.

1:06:03.140,1:06:06.320
If we have time in the end[br]we can get back to that.

1:06:06.320,1:06:09.590
But some qualifiers are again[br]relevant for some things,

1:06:09.590,1:06:13.180
start time, end time,[br]and others won't be.

1:06:13.180,1:06:16.280
Wikidata does try to offer you--

1:06:16.280,1:06:18.710
you may remember when I[br]clicked add qualifier,

1:06:18.710,1:06:22.170
it gave me kind of drop down[br]of some relevant qualifiers.

1:06:22.170,1:06:24.475
So it does try to[br]help you in that way.

1:06:27.280,1:06:28.160
Other question?

1:06:28.160,1:06:31.180
Are the values for[br]instance of already

1:06:31.180,1:06:33.310
mappable to external ontologies?

1:06:36.500,1:06:41.310
That is a complicated question.

1:06:41.310,1:06:43.490
I'll help people understand[br]the question first.

1:06:43.490,1:06:48.570
So an ontology is a[br]structure, some kind

1:06:48.570,1:06:52.350
of hierarchy or[br]cloud, of entities

1:06:52.350,1:06:54.510
and their interrelationships.

1:06:54.510,1:06:56.920
An ontology would[br]say, for example,

1:06:56.920,1:06:58.710
a person is a living thing.

1:06:58.710,1:06:59.670
So is a dog.

1:06:59.670,1:07:02.340
They're both living things,[br]but they're different things.

1:07:02.340,1:07:09.910
And then, you know, say[br]things about those entities

1:07:09.910,1:07:11.350
and their interrelationships.

1:07:11.350,1:07:13.300
Now there are many,[br]many competing,

1:07:13.300,1:07:17.230
or coexisting models[br]of ontology's.

1:07:17.230,1:07:19.840
Many of them were created[br]for specific needs.

1:07:19.840,1:07:25.170
Many of them want to be[br]a universal ontology.

1:07:25.170,1:07:27.790
But of course it's[br]impossible to quite

1:07:27.790,1:07:32.150
agree on one complete[br]and simple ontology.

1:07:32.150,1:07:34.240
And so there are[br]many ontology's.

1:07:34.240,1:07:38.520
Which brings up your question,[br]can we map across ontology's?

1:07:38.520,1:07:43.840
Can we say that when wiki data[br]says instance of book that

1:07:43.840,1:07:47.260
is equivalent to some other[br]ontology saying instance

1:07:47.260,1:07:49.940
of bibliographic record?

1:07:49.940,1:07:50.860
And the answer is yes.

1:07:50.860,1:07:52.360
There are some such mappings.

1:07:52.360,1:07:54.420
They are incomplete.

1:07:54.420,1:07:58.240
And there's no kind of[br]auto magic thing happening

1:07:58.240,1:08:01.180
in the wiki vis-a-vis[br]those other ontology's.

1:08:01.180,1:08:03.250
That's kind of[br]left as an exercise

1:08:03.250,1:08:06.280
for those dealing with those[br]other ontology's, and for tool

1:08:06.280,1:08:09.880
builders and other[br]platform improvements

1:08:09.880,1:08:13.050
beyond Wikidata itself.

1:08:13.050,1:08:13.750
OK.

1:08:13.750,1:08:15.190
Other questions?

1:08:15.190,1:08:17.430
Yeah, we have one from[br]the YouTube stream.

1:08:17.430,1:08:21.160
Someone asked, why can't I[br]link Howard Carter's occupation

1:08:21.160,1:08:26.439
to archeologists when I use[br]an info box that fetches info

1:08:26.439,1:08:28.960
from Wikidata?

1:08:28.960,1:08:33.160
Why can't I link it[br]from the info box?

1:08:33.160,1:08:35.500
So, someone on the[br]stream answered

1:08:35.500,1:08:37.659
saying, because it's[br]an improper connection,

1:08:37.659,1:08:39.700
because the target is not[br]about the subject only.

1:08:43.020,1:08:46.710
The target is not[br]about the subject?

1:08:46.710,1:08:48.479
If I understand the[br]question correctly,

1:08:48.479,1:08:53.130
what you would want to be able[br]to do is from within Wikipedia

1:08:53.130,1:08:59.130
be able to say occupation[br]and link to a Wikidata entry

1:08:59.130,1:09:01.050
about archeology.

1:09:01.050,1:09:03.569
That doesn't quite[br]work that way.

1:09:03.569,1:09:05.430
We will get to a[br]little discussion

1:09:05.430,1:09:08.460
of that in an upcoming[br]section of this talk.

1:09:08.460,1:09:13.260
So I will defer the rest[br]of my answer to then.

1:09:13.260,1:09:15.319
OK.

1:09:15.319,1:09:19.160
So we're done with[br]questions for this phase,

1:09:19.160,1:09:22.850
and my browser got[br]tired of waiting for me.

1:09:22.850,1:09:26.551
So, yes.

1:09:26.551,1:09:27.050
All right.

1:09:27.050,1:09:36.850
So we took a look at Wikidata,[br]and we took questions.

1:09:36.850,1:09:41.020
So now, let's teach[br]Wikidata some new things.

1:09:41.020,1:09:44.020
Some things it[br]doesn't already know.

1:09:44.020,1:09:47.109
Let's look at this item here.

1:09:47.109,1:09:50.950
So this item is about one[br]of my favorite writers,

1:09:50.950,1:09:53.840
an American writer[br]named Helen Dewitt.

1:09:53.840,1:10:01.570
Wikidata, of course, fondly[br]refers to her as q54674,

1:10:01.570,1:10:03.070
but we can call[br]her Helen Dewitt.

1:10:03.070,1:10:05.740
And what can we contribute here?

1:10:05.740,1:10:10.600
So Wikidata has far less[br]information about Helen Dewitt.

1:10:10.600,1:10:13.144
Most of you probably haven't[br]heard of her, that's OK.

1:10:13.144,1:10:14.560
What does Wikidata[br]know about her?

1:10:14.560,1:10:16.450
Well instance of human.

1:10:16.450,1:10:17.800
We have a photo of her.

1:10:17.800,1:10:18.780
She's female.

1:10:18.780,1:10:20.530
She's an American.

1:10:20.530,1:10:21.790
Her name is Helen.

1:10:21.790,1:10:22.630
Date of birth.

1:10:22.630,1:10:23.650
Place of birth.

1:10:23.650,1:10:25.970
She's an author, a[br]novelist, a writer.

1:10:25.970,1:10:28.840
She was educated at the[br]University of Oxford.

1:10:28.840,1:10:33.160
And Wikidata knows what[br]her official website is.

1:10:33.160,1:10:35.780
That's useful, but that's it.

1:10:35.780,1:10:37.780
Now we can contribute[br]information here.

1:10:37.780,1:10:43.120
For example, she's an American[br]author writing in English.

1:10:43.120,1:10:45.550
So we could add[br]that information.

1:10:45.550,1:10:48.430
We could click the[br]Add button here.

1:10:48.430,1:10:50.200
And this is a good[br]moment to acknowledge

1:10:50.200,1:10:54.830
that the user interface of[br]Wikidata is a work in progress.

1:10:54.830,1:10:56.740
It's not as intuitive[br]as it might be.

1:10:56.740,1:10:58.570
So you need to[br]understand that click--

1:10:58.570,1:11:01.630
to add a completely[br]new property,

1:11:01.630,1:11:04.060
You need to click[br]this Add button.

1:11:04.060,1:11:08.020
If you want to add an additional[br]value to the property official

1:11:08.020,1:11:11.530
website, you need to[br]click this Add button.

1:11:11.530,1:11:13.780
It makes a kind of[br]sense with a shaded box.

1:11:13.780,1:11:15.880
But, you know, you need[br]to kind of pay attention,

1:11:15.880,1:11:18.901
and it's not as[br]friendly as it might be.

1:11:18.901,1:11:20.650
[COUGHING] Excuse me.

1:11:20.650,1:11:23.380
So, let's add a property here.

1:11:23.380,1:11:25.690
Click the Add button.

1:11:25.690,1:11:29.740
Again, Wikidata tries to[br]be useful by suggesting

1:11:29.740,1:11:32.760
some relevant[br]properties for humans.

1:11:32.760,1:11:36.640
A bit more morbidly it suggests,[br]how about date of death?

1:11:36.640,1:11:38.700
That's not cool, Wikidata.

1:11:38.700,1:11:40.480
Helen Dewitt is still alive.

1:11:40.480,1:11:42.700
So I will not add[br]date of death, but I

1:11:42.700,1:11:46.140
can add languages spoken,[br]written, or signed.

1:11:46.140,1:11:48.370
OK, so I click that.

1:11:48.370,1:11:51.670
And she writes in English.

1:11:51.670,1:11:54.450
I just type English-- whoops.

1:11:54.450,1:11:56.750
Not in Hebrew.

1:11:56.750,1:11:58.380
Don't panic.

1:11:58.380,1:12:01.010
I type English here.

1:12:01.010,1:12:04.250
And, oh, and of course Wikidata[br]has auto-complete, right?

1:12:04.250,1:12:06.080
So it tries to help me along.

1:12:06.080,1:12:10.100
But you will notice that[br]it has all kinds of things

1:12:10.100,1:12:10.940
called English.

1:12:10.940,1:12:14.030
I mean, it turns out that[br]there is a place in Indiana

1:12:14.030,1:12:16.370
called English, Indiana.

1:12:16.370,1:12:17.150
Did I mean that?

1:12:17.150,1:12:20.210
No, of course I didn't mean[br]that she writes her books

1:12:20.210,1:12:21.961
in English, Indiana.

1:12:21.961,1:12:22.460
Right?

1:12:22.460,1:12:26.180
But, you know, Wikidata gives me[br]the option of linking to that.

1:12:26.180,1:12:30.530
I also don't mean the botanist[br]Carl Schwartz English.

1:12:30.530,1:12:32.870
No, no I mean the[br]west Germanic language

1:12:32.870,1:12:34.029
originating in England.

1:12:34.029,1:12:34.820
That's what I mean.

1:12:34.820,1:12:36.110
So I click that.

1:12:36.110,1:12:37.760
And I click Save.

1:12:37.760,1:12:38.450
And that's it.

1:12:38.450,1:12:41.780
Again I have just made[br]an edit to Wikidata.

1:12:41.780,1:12:47.750
I have just taught Wikidata[br]that this author speaks English.

1:12:47.750,1:12:50.370
Now, again, this[br]may be very obvious.

1:12:50.370,1:12:52.280
She's American.

1:12:52.280,1:12:54.560
Of course not all[br]Americans write in English.

1:12:54.560,1:12:56.930
It may be obvious if[br]you look at her books.

1:12:56.930,1:12:59.060
The important thing[br]is that now Wikidata

1:12:59.060,1:13:02.090
knows this as a piece of data.

1:13:02.090,1:13:04.610
And, again, think ahead[br]to queries, which we will

1:13:04.610,1:13:06.980
demonstrate in a little bit.

1:13:06.980,1:13:09.000
Without this piece[br]of information

1:13:09.000,1:13:14.060
that I just added, if I were to[br]ask Wikidata five minutes ago,

1:13:14.060,1:13:19.760
give me a list of novelists[br]writing in English, OK,

1:13:19.760,1:13:22.730
Wikidata would have returned[br]thousands of results.

1:13:22.730,1:13:27.600
But Helen Dewitt would[br]not have been among them.

1:13:27.600,1:13:32.000
Because up until two[br]minutes ago Wikidata

1:13:32.000,1:13:35.640
didn't know that Helen Dewitt[br]writes in English and not

1:13:35.640,1:13:37.520
in Spanish.

1:13:37.520,1:13:38.730
Do you see?

1:13:38.730,1:13:42.570
It is this explicit[br]statement that will now

1:13:42.570,1:13:46.560
make her be included in any[br]future queries that asks,

1:13:46.560,1:13:48.700
who are novelists[br]writing in English?

1:13:53.250,1:13:54.500
OK.

1:13:54.500,1:13:58.560
By the way, she's[br]a PhD in Classics.

1:13:58.560,1:14:05.590
She speaks-- or at least reads[br]and writes Latin and Greek,

1:14:05.590,1:14:07.270
ancient Greek, and I could--

1:14:07.270,1:14:09.610
I can-- I mean, I[br]happen to know that.

1:14:09.610,1:14:12.420
But wait, wait, wait,[br]wait, wait, you say.

1:14:12.420,1:14:14.130
What about original research?

1:14:14.130,1:14:18.890
I mean, you can't just add[br]stuff like that to Wikidata.

1:14:18.890,1:14:19.920
Don't you need sources?

1:14:19.920,1:14:22.860
Citations?

1:14:22.860,1:14:23.890
Of course I do.

1:14:23.890,1:14:25.020
Yes.

1:14:25.020,1:14:27.720
Let's add some sources to this.

1:14:27.720,1:14:31.410
So on Wikidata,[br]just like Wikipedia,

1:14:31.410,1:14:34.980
things should generally[br]be supported by citations,

1:14:34.980,1:14:36.990
by references.

1:14:36.990,1:14:43.290
And just like Wikipedia,[br]they aren't always supported

1:14:43.290,1:14:44.650
in that way.

1:14:44.650,1:14:48.870
OK so, I mean, I can[br]just add it to Wikidata.

1:14:48.870,1:14:49.442
Watch me.

1:14:49.442,1:14:50.400
I just did that, right?

1:14:50.400,1:14:54.450
I just added English and[br]Latin without any citation,

1:14:54.450,1:14:56.850
and I will not be[br]arrested for it.

1:14:56.850,1:14:59.520
Just like I could edit[br]a Wikipedia article

1:14:59.520,1:15:02.610
and add some information[br]without a citation.

1:15:02.610,1:15:03.600
It may stick.

1:15:03.600,1:15:06.810
It may stay in the article,[br]or it may be reverted.

1:15:06.810,1:15:11.010
It depends on the kind of[br]information I'm adding.

1:15:11.010,1:15:13.740
It depends how many people[br]are paying attention

1:15:13.740,1:15:15.060
to the article on Wikipedia.

1:15:15.060,1:15:18.420
And it works the[br]same way on Wikidata.

1:15:18.420,1:15:21.780
OK, so, you can add some[br]things without references.

1:15:21.780,1:15:23.970
Ideally, when you[br]add, information you

1:15:23.970,1:15:25.570
should include references.

1:15:25.570,1:15:30.990
So let's be good Wikidata[br]citizens and add a source.

1:15:30.990,1:15:34.395
Here is an article that[br]I prepared in advance.

1:15:38.100,1:15:39.370
This is Helen Dewitt.

1:15:39.370,1:15:44.450
And in this article,[br]somewhere, it actually

1:15:44.450,1:15:51.770
says right at the[br]bottom here, see,

1:15:51.770,1:15:54.990
Dewitt knows, in descending[br]order of proficiency, Latin,

1:15:54.990,1:15:57.010
ancient Greek, French,[br]German, Spanish,

1:15:57.010,1:15:59.460
and Portuguese, Dutch, Danish,[br]Norwegian, Swedish, Arabic,

1:15:59.460,1:16:01.680
Hebrew and Japanese.

1:16:01.680,1:16:04.770
This may sound[br]excessive, but it's true.

1:16:04.770,1:16:06.330
I met this woman.

1:16:06.330,1:16:09.670
So anyway, we don't have[br]to include all of that.

1:16:09.670,1:16:13.050
The point is this article from[br]a reasonably reliable source,

1:16:13.050,1:16:15.840
this magazine,[br]this interview, can

1:16:15.840,1:16:19.270
count as a source for[br]the languages she speaks.

1:16:19.270,1:16:20.700
So I copy the URL.

1:16:20.700,1:16:23.130
I just copied off my browser.

1:16:23.130,1:16:27.530
And, whoops-- that's not--

1:16:27.530,1:16:28.580
here we go.

1:16:28.580,1:16:31.610
And I can just add[br]a reference here

1:16:31.610,1:16:34.670
to the information that I[br]just added to Wikidata, right?

1:16:34.670,1:16:38.300
I can click Add Reference.

1:16:38.300,1:16:45.800
And then just say the reference[br]URL is, and I just paste.

1:16:45.800,1:16:48.840
I paste this URL.

1:16:48.840,1:16:50.160
Hit Enter.

1:16:50.160,1:16:51.060
And that's it.

1:16:51.060,1:16:55.380
And now the fact that she[br]speaks Latin has a reference.

1:16:55.380,1:16:58.320
If you look at the other[br]things here on Wikidata,

1:16:58.320,1:17:02.660
you can see that these IDs, for[br]example, have references, too.

1:17:02.660,1:17:03.420
Right?

1:17:03.420,1:17:06.570
In this case, the reference[br]just says, excuse me--

1:17:14.760,1:17:18.600
In this case it just as[br]imported from English Wikipedia.

1:17:18.600,1:17:24.970
But wait, you say, can[br]Wikipedia be a source?

1:17:24.970,1:17:26.620
Not properly, no.

1:17:26.620,1:17:30.100
I mean, just like Wikipedia[br]itself doesn't cite itself.

1:17:30.100,1:17:33.790
We don't say, this person[br]was born in this city

1:17:33.790,1:17:34.870
how do we know?

1:17:34.870,1:17:37.210
We read it on Wikipedia[br]in another language.

1:17:37.210,1:17:39.610
That's not a good citation.

1:17:39.610,1:17:41.400
It's not a good[br]citation for Wikidata

1:17:41.400,1:17:45.040
either so why do we put it here?

1:17:45.040,1:17:49.240
Well you can see the qualifier[br]here is different, right?

1:17:49.240,1:17:53.535
It's not reference URL, which[br]is what I put in for Latin here.

1:18:17.020,1:18:20.320
It's not reference URL here,[br]it's a different qualifier.

1:18:20.320,1:18:23.020
It says-- saying, imported from.

1:18:23.020,1:18:25.960
So this is not an[br]actual reference that

1:18:25.960,1:18:27.610
supports this piece of data.

1:18:27.610,1:18:30.730
It just shows where did[br]this data come from.

1:18:30.730,1:18:33.670
It's a slightly different[br]thing, because this data was

1:18:33.670,1:18:37.210
mass imported into Wikidata.

1:18:37.210,1:18:40.960
So it wasn't input by[br]hand by some volunteer.

1:18:40.960,1:18:44.770
It was imported into Wikidata[br]en masse by a script,

1:18:44.770,1:18:46.180
by a program.

1:18:46.180,1:18:49.820
And we want to know, where[br]did this number come from?

1:18:49.820,1:18:51.440
Well it came from[br]English Wikipedia.

1:18:51.440,1:18:54.130
So again, that's not[br]a proper reference

1:18:54.130,1:18:56.200
for the validity[br]of the information,

1:18:56.200,1:18:59.200
but it does at least tell us[br]it came from English Wikipedia.

1:18:59.200,1:19:03.460
We can click and look on[br]English Wikipedia and find out.

1:19:03.460,1:19:05.230
Maybe there's a[br]footnote there that

1:19:05.230,1:19:08.970
says where it did come from.

1:19:08.970,1:19:11.000
OK.

1:19:11.000,1:19:15.320
So this was an example of[br]teaching Wikidata something

1:19:15.320,1:19:16.910
that it didn't know.

1:19:16.910,1:19:18.512
Something about the languages.

1:19:18.512,1:19:20.720
And of course I could add[br]this reference for English.

1:19:20.720,1:19:23.210
I could add all the other[br]languages that she speaks.

1:19:23.210,1:19:26.060
And I won't bore you with[br]that, but that is basically

1:19:26.060,1:19:27.050
how it's done.

1:19:27.050,1:19:29.720
So you click this Add to[br]add a completely new--

1:19:32.650,1:19:34.030
completely new statement.

1:19:34.030,1:19:36.250
Now, by the way, the fact[br]that these are the only two

1:19:36.250,1:19:39.220
suggestions that[br]Wikidata can think of,

1:19:39.220,1:19:42.100
doesn't mean these[br]are the only options.

1:19:42.100,1:19:46.750
OK, you can just type[br]anything that may be relevant.

1:19:46.750,1:19:50.950
We could add, for[br]example, award.

1:19:50.950,1:19:52.570
Just start typing award.

1:19:52.570,1:19:54.910
And here I have I have[br]a bunch of properties

1:19:54.910,1:19:56.510
that are relevant for awards.

1:19:56.510,1:20:00.100
Awards received, together[br]with, conferred by, right?

1:20:00.100,1:20:05.790
There's all kinds of properties[br]that I could rely on.

1:20:05.790,1:20:09.600
And of course there is a list of[br]all the properties of Wikidata.

1:20:09.600,1:20:11.580
And that list is[br]also sorted by type.

1:20:11.580,1:20:15.480
So yes, there is a list of[br]properties relevant to people

1:20:15.480,1:20:17.130
so that you don't have to guess.

1:20:17.130,1:20:18.660
But a surprising[br]amount of the time

1:20:18.660,1:20:22.760
you can just start typing[br]and get the right properties

1:20:22.760,1:20:25.340
suggested to you.

1:20:25.340,1:20:27.230
OK.

1:20:27.230,1:20:33.050
So we taught Wikidata[br]something new,

1:20:33.050,1:20:38.980
and now let's teach Wikidata[br]something completely new.

1:20:38.980,1:20:39.480
Right?

1:20:39.480,1:20:42.480
So how do we create[br]a new Wikidata item?

1:20:42.480,1:20:46.880
So, like I said, if I[br]created a Wikipedia article

1:20:46.880,1:20:49.520
about something that was[br]not previously covered

1:20:49.520,1:20:53.540
on any other[br]Wikipedia, chances are

1:20:53.540,1:20:57.170
there would not be an already[br]existing Wikidata item.

1:20:57.170,1:21:03.190
Sometimes there might[br]be, because Wikidata

1:21:03.190,1:21:06.857
does have 25 million entities.

1:21:06.857,1:21:08.190
But sometimes there wouldn't be.

1:21:08.190,1:21:10.148
So, first of all, I could[br]search for it, right?

1:21:10.148,1:21:14.210
So I could go to Wikidata[br]to the search box

1:21:14.210,1:21:17.390
here and just start typing, and[br]search for what I want, right?

1:21:17.390,1:21:20.690
So if I'm searching for Helen[br]Dewitt I just say Helen,

1:21:20.690,1:21:25.590
and I can see whether[br]or not it exists.

1:21:25.590,1:21:29.240
And there's a detailed search[br]results page, et cetera,

1:21:29.240,1:21:33.074
where I can where I can find out[br]if the item does exist or not.

1:21:33.074,1:21:35.240
Excuse me, this reminds me[br]of a very important thing

1:21:35.240,1:21:36.620
I wanted to[br]demonstrate, and that

1:21:36.620,1:21:42.710
is the multilingualism[br]of Wikidata.

1:21:42.710,1:21:49.340
So remember all these[br]labels in other languages.

1:21:49.340,1:21:54.390
Wikidata knows what to call[br]Helen Dewitt in Hebrew.

1:21:54.390,1:22:00.800
And it will show it to Wikidata[br]users whose language is Hebrew.

1:22:00.800,1:22:04.220
Mine is set to[br]English, for your sake.

1:22:04.220,1:22:08.830
But if I change this I go to[br]Preferences here and change

1:22:08.830,1:22:09.740
my language.

1:22:09.740,1:22:15.475
[INAUDIBLE] All[br]right, and I hit Save.

1:22:15.475,1:22:20.350
Wikidata will start[br]talking to me in Hebrew.

1:22:20.350,1:22:23.090
Now brace yourselves.

1:22:23.090,1:22:24.620
Are you ready?

1:22:24.620,1:22:28.430
Don't panic, it's right to left.

1:22:28.430,1:22:32.630
Oh my god everything[br]is topsy-turvy.

1:22:32.630,1:22:36.590
So this is the same[br]article in Hebrew.

1:22:36.590,1:22:39.290
So the sidebar has[br]switched direction,

1:22:39.290,1:22:41.300
and I know most of[br]you cannot read it.

1:22:41.300,1:22:42.480
Bear with me.

1:22:42.480,1:22:44.750
This is the label[br]that we previously

1:22:44.750,1:22:46.840
saw in the label box.

1:22:46.840,1:22:49.580
This is how you spell[br]Helen Dewitt in Hebrew.

1:22:49.580,1:22:52.550
And here is the[br]description in Hebrew.

1:22:52.550,1:22:54.980
It's not the description in[br]English, this description,

1:22:54.980,1:22:57.380
American writer, which[br]I was shown previously.

1:22:57.380,1:23:00.740
Now I'm shown the Hebrew[br]description, appropriately.

1:23:00.740,1:23:03.500
But more interestingly,[br]oh my god!

1:23:03.500,1:23:07.640
All these statements[br]are suddenly in Hebrew.

1:23:07.640,1:23:08.940
How did that happen?

1:23:11.570,1:23:15.560
Well this tiny word here[br]is the very concise way

1:23:15.560,1:23:22.450
to say in Hebrew, instance of,[br]and this word here means human.

1:23:22.450,1:23:25.960
So these are links to[br]the same things, right?

1:23:25.960,1:23:28.100
It still links to Q5.

1:23:28.100,1:23:31.780
Q5 is the Wikidata[br]entity for human.

1:23:31.780,1:23:33.370
These are still the same things.

1:23:33.370,1:23:37.600
But because Wikidata has[br]multiple labels for everything,

1:23:37.600,1:23:39.580
it has multiple[br]labels for items.

1:23:39.580,1:23:42.760
And it also has multiple[br]labels for property names.

1:23:42.760,1:23:46.450
So Wikidata knows how[br]to say, instance of,

1:23:46.450,1:23:50.140
and award received,[br]in other languages.

1:23:50.140,1:23:54.490
That is why it is able to show[br]me all this data in Hebrew

1:23:54.490,1:23:59.890
even if none of that data was[br]actually input into Wikidata

1:23:59.890,1:24:01.870
by a Hebrew speaker.

1:24:01.870,1:24:04.900
That data could have been[br]input by English speakers,

1:24:04.900,1:24:08.230
but thanks to the[br]fact that someone once

1:24:08.230,1:24:12.760
translated the word[br]photo into Hebrew,

1:24:12.760,1:24:14.830
I can see this field in Hebrew.

1:24:17.750,1:24:21.230
So one of the things you[br]can do to help Wikidata,

1:24:21.230,1:24:23.600
right now, without[br]any special knowledge

1:24:23.600,1:24:26.210
is to help translate[br]those labels.

1:24:26.210,1:24:29.030
Every label only needs to[br]be translated just once.

1:24:29.030,1:24:31.310
So you can see that all[br]of these properties, date

1:24:31.310,1:24:34.720
of birth, name et cetera,[br]they all have Hebrew labels.

1:24:34.720,1:24:36.760
Maybe one of these would not.

1:24:36.760,1:24:38.361
No, they all have Hebrew labels.

1:24:38.361,1:24:39.110
Doing pretty good.

1:24:42.960,1:24:45.810
And I'm able to search[br]in my own language.

1:24:45.810,1:24:48.210
I'm able to click Add.

1:24:48.210,1:24:49.890
This word is Add,[br]so I click this,

1:24:49.890,1:24:51.780
and now I have the Add screen.

1:24:51.780,1:24:55.860
It all speaks my language,[br]and it's awesome.

1:24:55.860,1:25:00.330
And now for your sake I[br]will switch back to English,

1:25:00.330,1:25:03.090
but it is important[br]to know you can

1:25:03.090,1:25:05.740
edit Wikidata in any language.

1:25:05.740,1:25:09.050
And it is far more multi-lingual[br]and multi-lingual friendly

1:25:09.050,1:25:13.260
than, for example commons, which[br]is also a project we all share.

1:25:13.260,1:25:17.730
But commons has some limitations[br]on how multi-lingual it is.

1:25:17.730,1:25:21.410
For example, the category[br]names, et cetera.

1:25:21.410,1:25:23.270
OK.

1:25:23.270,1:25:25.670
So we were beginning[br]to discuss creating

1:25:25.670,1:25:27.140
something completely new.

1:25:27.140,1:25:29.360
AUDIENCE: Quick[br]questions, if that's OK?

1:25:29.360,1:25:30.980
So there's two questions on IRC.

1:25:30.980,1:25:33.890
The first one is, can you[br]show search for something

1:25:33.890,1:25:35.420
like getting the list of things?

1:25:35.420,1:25:38.360
I want to learn how to search[br]for something properly like,

1:25:38.360,1:25:43.705
show me all the items with[br]this value of this property.

1:25:43.705,1:25:45.080
ASAF BARTOV: Yes.

1:25:45.080,1:25:47.540
That is part of[br]this talk, but I'll

1:25:47.540,1:25:49.250
get to that in a[br]little bit later.

1:25:49.250,1:25:52.010
There's a whole section where I[br]will demonstrate the very, very

1:25:52.010,1:25:55.190
powerful query[br]system of Wikidata

1:25:55.190,1:25:57.170
where I will cash[br]that check that I gave

1:25:57.170,1:25:59.090
at the beginning of[br]all these painters

1:25:59.090,1:26:01.029
who are sons of painters[br]queries et cetera

1:26:01.029,1:26:02.570
So I will demonstrate[br]how to do that.

1:26:02.570,1:26:04.190
AUDIENCE: Other question.

1:26:04.190,1:26:07.250
How does Wikidata data deal[br]with link rot, and other issues

1:26:07.250,1:26:09.680
streaming from their URL refs.

1:26:13.528,1:26:16.290
ASAF BARTOV: URLs break.

1:26:16.290,1:26:18.730
We call that link rot.

1:26:18.730,1:26:22.470
Wikidata doesn't have[br]any particular magic

1:26:22.470,1:26:24.730
around link rot,[br]just like Wikipedia.

1:26:24.730,1:26:29.100
So if you do use a bare[br]URL it may well rot.

1:26:29.100,1:26:34.230
But you can add qualifiers[br]with back up URLs else

1:26:34.230,1:26:37.680
on the Internet Archive, or[br]another mirroring service.

1:26:37.680,1:26:42.780
And potentially that could be[br]a software feature for Wikidata

1:26:42.780,1:26:46.590
to automatically save[br]or ensure that something

1:26:46.590,1:26:48.660
is saved on Internet[br]Archive, but I don't

1:26:48.660,1:26:50.670
know that it is doing so now.

1:26:50.670,1:26:56.040
So, just like Wikipedia, if[br]it is a bear URL it may rot.

1:26:56.040,1:27:00.240
And may need to be[br]replaced, possibly by bot.

1:27:00.240,1:27:01.390
Other questions?

1:27:09.840,1:27:12.650
All right, so let's[br]talk about how you

1:27:12.650,1:27:15.090
create a completely new item.

1:27:15.090,1:27:16.300
It's very simple.

1:27:16.300,1:27:21.810
You go to Wikidata and you[br]click here on the side.

1:27:21.810,1:27:30.180
There's a link, create new item,[br]which gives you this screen.

1:27:30.180,1:27:35.030
And let's create an[br]item about a book

1:27:35.030,1:27:39.500
that I'm reading right now[br]by this Bulgarian writer.

1:27:39.500,1:27:43.950
So we have an article about this[br]writer guy named Deyan Enev.

1:27:43.950,1:27:48.530
But we don't have an[br]article or a Wikidata item

1:27:48.530,1:28:07.980
about one of his famous[br]books called Circus Bulgaria.

1:28:07.980,1:28:10.050
That's the book I'm reading,[br]his first collection

1:28:10.050,1:28:11.216
of short stories in English.

1:28:11.216,1:28:14.280
Circus Bulgaria came out[br]in 2010, Portobello Books,

1:28:14.280,1:28:17.099
translated by Kapka Kassabova.

1:28:17.099,1:28:18.390
So that's the book I'm reading.

1:28:18.390,1:28:20.520
As you can see it's not[br]a link on Wikipedia.

1:28:20.520,1:28:23.370
There's no article about[br]it, and there's not even

1:28:23.370,1:28:26.310
a Wikidata entity item about it.

1:28:26.310,1:28:32.220
But we can totally create[br]it, even without a Wikipedia

1:28:32.220,1:28:33.090
article.

1:28:33.090,1:28:34.980
So let's create this new item.

1:28:34.980,1:28:37.260
Let's create it in[br]English for the purposes

1:28:37.260,1:28:38.880
of our demonstration.

1:28:38.880,1:28:44.910
The name of the item[br]is Circus Bulgaria.

1:28:44.910,1:28:47.520
Circus Bulgaria,[br]that's the name.

1:28:47.520,1:28:50.670
Not Circus Bulgaria[br]parentheses book,

1:28:50.670,1:28:53.520
or anything you may be[br]used to from Wikipedia.

1:28:53.520,1:28:56.520
It's the actual[br]name of the book,

1:28:56.520,1:29:00.450
and the description,[br]again, remember,

1:29:00.450,1:29:03.270
the description field[br]is just to kind of help

1:29:03.270,1:29:08.681
tell apart this Circus Bulgaria[br]from any other potential Circus

1:29:08.681,1:29:09.180
Bulgaria.

1:29:09.180,1:29:11.280
Maybe there's a[br]film or something.

1:29:11.280,1:29:20.480
So it's enough to just say[br]something like short story

1:29:20.480,1:29:23.270
collection.

1:29:23.270,1:29:27.830
I might add by Deyan Enev[br]and if just in case, again,

1:29:27.830,1:29:31.910
some future other short story[br]collection by some other author

1:29:31.910,1:29:33.560
happens to have that same name.

1:29:33.560,1:29:36.391
That should be[br]disambiguating enough.

1:29:36.391,1:29:36.890
OK.

1:29:36.890,1:29:39.770
Short story collection[br]by Deyan Enev.

1:29:39.770,1:29:42.050
I could have aliases for this.

1:29:42.050,1:29:47.240
The aliases assist find-ability.

1:29:47.240,1:29:51.020
This particular book has just[br]this one name, so that's fine.

1:29:51.020,1:29:52.260
And I click Create.

1:29:52.260,1:29:52.760
That's it.

1:29:52.760,1:29:55.990
I just start with a[br]label, and a description.

1:29:55.990,1:29:58.740
I click Create.

1:29:58.740,1:30:03.890
I have a brand new queue number[br]for my new Wikidata item.

1:30:03.890,1:30:05.960
And Wikidata knows[br]what to call it.

1:30:05.960,1:30:09.320
And a description in[br]one language at least.

1:30:09.320,1:30:11.930
And that's it, and I[br]can start populating it.

1:30:11.930,1:30:15.050
As it can see, it it[br]has no site links,

1:30:15.050,1:30:17.450
but it's ready to be taught.

1:30:17.450,1:30:20.450
So, for example, I[br]can start by teaching

1:30:20.450,1:30:24.610
it the name of the book[br]in another language

1:30:24.610,1:30:25.870
that I happened to speak.

1:30:29.050,1:30:31.720
Now it has two labels[br]in English and Hebrew.

1:30:31.720,1:30:36.880
I could also look[br]up the book Areon,

1:30:36.880,1:30:39.510
the original Bulgarian[br]label for this book.

1:30:39.510,1:30:41.550
Seems relevant.

1:30:41.550,1:30:43.320
Again, I do not speak Bulgarian.

1:30:43.320,1:30:49.860
But I can go to the Bulgarian[br]Wikipedia through into Wiki.

1:30:49.860,1:30:51.510
This is this gentleman.

1:30:51.510,1:30:54.510
And I could find--

1:30:54.510,1:30:59.190
I can read Cyrillic so[br]I could easily find--

1:30:59.190,1:31:00.030
when I say easily--

1:31:02.940,1:31:05.710
when I say easily--

1:31:05.710,1:31:12.731
maybe not so easy, but[br]I can search for it.

1:31:21.070,1:31:22.180
Here we go.

1:31:22.180,1:31:25.190
Tsirk Bulgaria.

1:31:25.190,1:31:27.510
That is the name of the book.

1:31:27.510,1:31:28.910
Tsirk, as in circus.

1:31:28.910,1:31:30.440
No problem.

1:31:30.440,1:31:32.725
So I just copy this right here.

1:31:35.240,1:31:38.090
And I go back to my new item.

1:31:38.090,1:31:45.725
My new item, which is here,[br]and I edit the Bulgarian field.

1:31:48.260,1:31:49.950
And here it is.

1:31:49.950,1:31:50.720
Awesome.

1:31:50.720,1:31:51.220
All right.

1:31:51.220,1:31:55.420
But I still haven't told[br]Wikidata anything about this.

1:31:55.420,1:31:56.920
I know I'm talking about a book.

1:31:56.920,1:31:59.110
Wikidata that doesn't[br]know that yet.

1:31:59.110,1:32:02.630
So let's start by[br]adding some statements.

1:32:02.630,1:32:05.390
First of all, I click Add.

1:32:05.390,1:32:07.190
Wikidata sensibly[br]says, how about we

1:32:07.190,1:32:08.630
start with instance of.

1:32:08.630,1:32:11.090
Tell me what kind of animal--[br]no, not kind of animal.

1:32:11.090,1:32:13.940
What kind of thing are you[br]trying to describe here?

1:32:13.940,1:32:18.130
Well it's an instance of a book.

1:32:18.130,1:32:20.930
Not in Hebrew, please.

1:32:20.930,1:32:22.180
So it's an instance of a book.

1:32:22.180,1:32:23.763
I could even be a[br]little more specific

1:32:23.763,1:32:31.920
and say it's an instance of[br]a short story collection.

1:32:31.920,1:32:34.620
There we go, short[br]story collection.

1:32:34.620,1:32:36.800
I hit Save.

1:32:36.800,1:32:37.430
Awesome.

1:32:37.430,1:32:39.680
So now we know what[br]kind of thing it is.

1:32:39.680,1:32:42.860
It's not a human, it's not a[br]mountain, it's not a concept.

1:32:42.860,1:32:44.760
It's a short story collection.

1:32:44.760,1:32:46.400
Now I can add some other things.

1:32:46.400,1:32:48.770
See, Wikidata is[br]already working for me.

1:32:48.770,1:32:51.020
Because it's a short[br]story collection

1:32:51.020,1:32:53.960
it's offering me to populate[br]these properties, and not

1:32:53.960,1:32:54.890
other ones.

1:32:54.890,1:32:56.990
Publication date,[br]original language,

1:32:56.990,1:33:00.350
genre, country of origin,[br]these are all relevant, right?

1:33:00.350,1:33:04.220
So let's start with original[br]language of the work

1:33:04.220,1:33:07.410
is Bulgarian.

1:33:07.410,1:33:09.810
Not Bulgaria, Bulgarian.

1:33:09.810,1:33:12.040
This is the item I want to link.

1:33:12.040,1:33:21.570
Hit Save, and whatever.

1:33:21.570,1:33:22.890
Author.

1:33:22.890,1:33:26.540
Let's identify the author.

1:33:26.540,1:33:29.350
So the author, the main[br]creator of the work,

1:33:29.350,1:33:32.470
is that gentleman Deyan Enev.

1:33:32.470,1:33:34.750
And remember, he has[br]a Wikipedia article.

1:33:34.750,1:33:37.210
He also has a Wikidata entity.

1:33:37.210,1:33:39.640
So Wikidata does know about him.

1:33:39.640,1:33:48.930
So I hit Save, and I can add[br]something about the translator.

1:33:52.530,1:33:54.390
And what was that lady's name?

1:33:57.990,1:34:00.120
Kapka Kassabova.

1:34:00.120,1:34:05.430
Now it so happens that Wikidata[br]already knows about this lady.

1:34:08.330,1:34:08.840
See?

1:34:08.840,1:34:12.290
So I can just start typing[br]and then just link to it.

1:34:12.290,1:34:12.840
Awesome.

1:34:12.840,1:34:13.824
But what if it didn't?

1:34:13.824,1:34:15.740
What if it was translated[br]by someone who isn't

1:34:15.740,1:34:17.690
already covered on Wikidata?

1:34:17.690,1:34:22.190
Well I could just type[br]the name as a string,

1:34:22.190,1:34:25.760
but ideally I could[br]create a Wikidata entity

1:34:25.760,1:34:28.940
about this translator so[br]that there is a possibility

1:34:28.940,1:34:30.350
to link to her.

1:34:33.560,1:34:36.920
Now I might actually[br]add a qualifier here

1:34:36.920,1:34:40.310
because, she's not the[br]translator of the book, right?

1:34:40.310,1:34:43.620
She's the translator of[br]the book into English.

1:34:43.620,1:34:44.440
Right.

1:34:44.440,1:34:50.151
So the language that she[br]translated into is English.

1:34:50.151,1:34:50.650
Right?

1:34:50.650,1:34:53.620
This book-- remember[br]I'm describing the book.

1:34:53.620,1:34:55.376
The item is about the book.

1:34:55.376,1:34:57.250
So the book would have[br]a different translator

1:34:57.250,1:34:58.510
into Polish.

1:34:58.510,1:35:02.320
So this is an example of[br]a property or a statement

1:35:02.320,1:35:06.430
that doesn't make sense without[br]one of those qualifiers.

1:35:06.430,1:35:08.140
It's just not correct.

1:35:08.140,1:35:11.320
It doesn't make sense to[br]say that translator is.

1:35:11.320,1:35:14.950
The English translator, or[br]even this English translator.

1:35:14.950,1:35:17.770
In 50 years maybe there would[br]be an additional English

1:35:17.770,1:35:18.940
translation.

1:35:18.940,1:35:24.774
So that's an example of[br]needing that qualifier.

1:35:24.774,1:35:27.190
And of course I could go on[br]and populate the other fields.

1:35:27.190,1:35:29.710
We don't have to[br]do that right now.

1:35:29.710,1:35:32.960
Publication date, country[br]of origin, et cetera.

1:35:32.960,1:35:35.440
So this is already beginning[br]to look like all those items

1:35:35.440,1:35:38.440
that we already saw, but just[br]a moment ago it didn't exist.

1:35:38.440,1:35:43.920
Just a moment ago Wikidata[br]had no concept of this work.

1:35:43.920,1:35:46.500
This happens to be one[br]of his notable works.

1:35:46.500,1:35:52.080
So I could actually go to the[br]item about Deyan Enev which

1:35:52.080,1:35:56.190
has all this information[br]already, occupation, languages,

1:35:56.190,1:35:59.170
and add a property.

1:35:59.170,1:36:01.050
Remember, I'm not[br]limited to these.

1:36:01.050,1:36:06.180
I can add a property[br]called notable works,

1:36:06.180,1:36:08.670
and mention my new item.

1:36:08.670,1:36:12.120
Circus Bulgaria.

1:36:12.120,1:36:12.750
See?

1:36:12.750,1:36:15.180
My new item is[br]showing up, and thanks

1:36:15.180,1:36:18.660
to this description that I[br]wrote, short story collection,

1:36:18.660,1:36:22.650
it's already appearing here in[br]the dropdown very conveniently.

1:36:22.650,1:36:24.270
So I linked to this.

1:36:24.270,1:36:25.154
I hit Save.

1:36:28.680,1:36:32.310
Ideally again I should find[br]some references showing

1:36:32.310,1:36:34.620
that this is a[br]notable work by him,

1:36:34.620,1:36:37.000
but we won't spend[br]time on that right now.

1:36:37.000,1:36:39.010
But the point is we[br]created a new item.

1:36:39.010,1:36:40.410
We populated it a little bit.

1:36:40.410,1:36:44.400
We linked to it so that it's[br]more discoverable by mentioning

1:36:44.400,1:36:47.760
it in the author name, and[br]of course the book item

1:36:47.760,1:36:50.710
itself mentions the author[br]and links to the author.

1:36:50.710,1:36:52.770
So that's all good.

1:36:52.770,1:36:57.780
One last thing we shall do is[br]give it some useful identifier

1:36:57.780,1:37:02.880
so let's add, say, the[br]Library of Congress record

1:37:02.880,1:37:03.940
for this book.

1:37:03.940,1:37:04.440
OK.

1:37:04.440,1:37:07.710
So I have prepared[br]this in advance.

1:37:07.710,1:37:08.760
Ooh.

1:37:08.760,1:37:12.720
Just in time, with 80 seconds to[br]go before it's giving up on me.

1:37:12.720,1:37:14.310
Oh it has already[br]given up on me.

1:37:14.310,1:37:15.490
That is very unfortunate.

1:37:23.300,1:37:29.110
So I go to the Library of[br]Congress and I find this book.

1:37:29.110,1:37:33.050
I find this entry, right?

1:37:33.050,1:37:37.320
In the Library of Congress[br]database about this book.

1:37:37.320,1:37:39.120
And it has a permalink.

1:37:39.120,1:37:42.570
It has a kind of guaranteed[br]to be permanent link.

1:37:42.570,1:37:47.950
I can just copy that link,[br]go back to my little book,

1:37:47.950,1:37:55.770
and say the Library of Congress.

1:37:55.770,1:38:01.070
Yeah, LCCN, that's what they[br]call their IDs, the call

1:38:01.070,1:38:02.120
number.

1:38:02.120,1:38:06.502
And I paste it here.

1:38:06.502,1:38:08.210
I actually don't need the URL.

1:38:08.210,1:38:09.136
I need just a number.

1:38:12.440,1:38:13.520
And there we go.

1:38:13.520,1:38:16.550
I have added it,[br]and now Wikidata

1:38:16.550,1:38:20.630
knows how to find bibliographic[br]information about this book.

1:38:20.630,1:38:24.710
And any re-user of[br]Wikidata, some program,

1:38:24.710,1:38:28.950
some tool that connects[br]books to authors

1:38:28.950,1:38:32.870
or does statistical analysis or[br]whatever, some future yet to be

1:38:32.870,1:38:35.090
imagined tool[br]could automatically

1:38:35.090,1:38:39.170
find additional metadata on the[br]Library of Congress site thanks

1:38:39.170,1:38:41.840
to this connection[br]that I just made.

1:38:41.840,1:38:44.150
And of course I could[br]add many other IDs

1:38:44.150,1:38:46.460
to other catalogs[br]around the world,

1:38:46.460,1:38:48.150
and we won't do that right now.

1:38:48.150,1:38:51.840
You can see that it's now[br]showing up under identifiers.

1:38:51.840,1:38:56.330
So this is how we created[br]a brand new piece of data.

1:38:56.330,1:38:59.632
Questions about this,[br]about creating new items?

1:39:18.100,1:39:19.180
Yeah, all right.

1:39:19.180,1:39:25.510
So we've seen how to contribute[br]to Wikidata on our own,

1:39:25.510,1:39:26.350
kind of through--

1:39:26.350,1:39:27.840
directly through Wikidata.

1:39:30.680,1:39:35.220
Now you may you may be[br]thinking, but Asaf, this

1:39:35.220,1:39:39.880
sounds like a ton[br]of work recording

1:39:39.880,1:39:44.500
all of these little tiny bits of[br]information about every person

1:39:44.500,1:39:47.410
and every book and every town.

1:39:47.410,1:39:50.520
And if you think that[br]you would be correct.

1:39:50.520,1:39:52.730
That is a ton of work.

1:39:52.730,1:39:54.600
It's a lot of work.

1:39:54.600,1:39:59.930
However, it is centralized, so[br]it is reusable on other wikis

1:39:59.930,1:40:03.860
and we will show in just a[br]moment how we pull information

1:40:03.860,1:40:07.296
from Wikidata into[br]Wikipedia or other projects.

1:40:10.860,1:40:13.780
We will show that[br]in just a moment.

1:40:13.780,1:40:18.660
But here's an[br]awesome little game

1:40:18.660,1:40:23.205
that we Wikidata[br]volunteer, Magnis Monska,

1:40:23.205,1:40:30.900
has authored called the[br]Wikidata game, in which he

1:40:30.900,1:40:31.920
tricks people--

1:40:31.920,1:40:35.730
sorry, helps people[br]make contributions

1:40:35.730,1:40:41.500
to Wikidata in a very,[br]very easy and pleasant way.

1:40:41.500,1:40:44.410
Let's look at the Wikidata game.

1:40:44.410,1:40:47.840
So the first thing you need[br]to do in that Wikidata game

1:40:47.840,1:40:50.660
is to log in,[br]because the Wikidata

1:40:50.660,1:40:53.150
game makes edits in your name.

1:40:53.150,1:40:54.980
So we need to authorize it.

1:40:54.980,1:40:57.250
It's perfectly safe.

1:40:57.250,1:41:01.090
And after you do that you[br]can go to the Wikidata game.

1:41:01.090,1:41:02.020
So this is the game.

1:41:02.020,1:41:03.520
Now I'm logged in.

1:41:03.520,1:41:05.230
And the Wikidata game[br]actually includes

1:41:05.230,1:41:06.970
a number of different games.

1:41:06.970,1:41:09.310
Let's start with a person game.

1:41:09.310,1:41:14.170
So Wikidata shows you--

1:41:14.170,1:41:20.800
shows you an item, and asks[br]you a very simple question.

1:41:20.800,1:41:23.200
Person, or not a person?

1:41:26.410,1:41:30.550
So Wikidata goes through[br]Wikidata entities

1:41:30.550,1:41:35.540
that don't even have the[br]instance of property.

1:41:35.540,1:41:37.520
Which is why Wikidata[br]doesn't know,

1:41:37.520,1:41:41.120
literally doesn't know, if this[br]is a person, or a mountain,

1:41:41.120,1:41:44.390
or a city, or a country,[br]or anything else.

1:41:44.390,1:41:47.150
So it asks you, because this[br]is the kind of question that

1:41:47.150,1:41:50.300
Wikidata cannot[br]decide on its own,

1:41:50.300,1:41:54.800
but for us humans it's generally[br]trivial to be able to say

1:41:54.800,1:41:58.220
whether something that we're[br]looking at is a person or not.

1:41:58.220,1:42:03.590
It gets slightly trickier when[br]the information is in Javanese,

1:42:03.590,1:42:06.470
as it is here,[br]rather than English.

1:42:06.470,1:42:10.010
So this item happens to[br]be described in Javanese.

1:42:10.010,1:42:14.360
My Javanese, spoken in[br]Indonesia, is very weak.

1:42:14.360,1:42:19.620
However, I can tell that[br]this is not a person.

1:42:19.620,1:42:20.730
How can I tell?

1:42:20.730,1:42:23.220
Without understanding[br]a word of Japanese

1:42:23.220,1:42:25.950
I see that it mentions[br]1000 kilometers

1:42:25.950,1:42:28.860
and square kilometers, see?

1:42:28.860,1:42:32.520
So this is about a[br]place, or an area,

1:42:32.520,1:42:36.090
or a region, or whatever,[br]but not a person.

1:42:36.090,1:42:39.060
So this is an[br]example of how even

1:42:39.060,1:42:41.100
without understanding[br]language you can sometimes

1:42:41.100,1:42:42.400
make a determination.

1:42:42.400,1:42:45.030
However, of course,[br]you should be sure.

1:42:45.030,1:42:47.700
This is definitely not[br]what the Wikipedia article

1:42:47.700,1:42:49.150
about a person looks like.

1:42:49.150,1:42:50.430
So this is not a person.

1:42:50.430,1:42:52.780
I just click it and I'm[br]shown the next item.

1:42:56.600,1:42:59.660
This item is in another[br]language I do not speak,

1:42:59.660,1:43:00.950
and I just don't know.

1:43:00.950,1:43:03.740
I do not know if this is[br]about a person or not.

1:43:03.740,1:43:07.350
So I click Not Sure.

1:43:07.350,1:43:11.190
This is in Swedish, and[br]it's about Sulawesi, still

1:43:11.190,1:43:13.770
Indonesia.

1:43:13.770,1:43:16.530
And it is not about a person.

1:43:16.530,1:43:18.150
I have enough Swedish for that.

1:43:18.150,1:43:21.750
So I click not a person.

1:43:21.750,1:43:24.420
Now, you may say,[br]well, do I really

1:43:24.420,1:43:28.350
have to deal with all these[br]languages that I don't speak?

1:43:28.350,1:43:29.190
The answer is no.

1:43:29.190,1:43:30.630
You don't have to.

1:43:30.630,1:43:32.580
Here at the bottom[br]of the Wikidata game

1:43:32.580,1:43:33.840
there are settings.

1:43:33.840,1:43:38.270
You can click that[br]and tell Wikidata,

1:43:38.270,1:43:41.840
I cannot even read[br]Chinese or Japanese,

1:43:41.840,1:43:44.600
so please don't show me[br]items in those languages.

1:43:44.600,1:43:47.060
Because I wouldn't[br]even be able to guess.

1:43:47.060,1:43:50.000
I prefer these languages in[br]which I can relatively easily

1:43:50.000,1:43:51.380
make determinations.

1:43:51.380,1:43:54.601
And I can even tell Wikidata to[br]only show me these languages.

1:43:54.601,1:43:55.100
You see?

1:43:55.100,1:43:57.350
This was not selected,[br]which is why I

1:43:57.350,1:44:00.600
was shown some other languages.

1:44:00.600,1:44:04.240
I could say, only use[br]these languages, and save.

1:44:04.240,1:44:06.100
And now I can try[br]this game again.

1:44:06.100,1:44:07.980
However, that can[br]slow it down a little.

1:44:07.980,1:44:09.000
So here we go.

1:44:09.000,1:44:11.640
Here's a Spanish-- which[br]is one of the languages I

1:44:11.640,1:44:14.640
told Wikidata game it can use.

1:44:14.640,1:44:16.480
This is a Spanish item.

1:44:16.480,1:44:19.265
Now is it about a person or not?

1:44:22.120,1:44:23.230
It is not about a person.

1:44:25.906,1:44:26.780
Is it about a person?

1:44:29.155,1:44:29.655
No.

1:44:32.900,1:44:35.180
Yes, it is right?

1:44:35.180,1:44:38.550
Monk Cistercian, Pedro[br]de Ovideo Falconi.

1:44:38.550,1:44:40.890
That sounds like a person.

1:44:40.890,1:44:42.680
Frau Pedro Nasser.

1:44:42.680,1:44:44.960
Yeah, he was born[br]in Madrid 1577.

1:44:44.960,1:44:46.280
This is a person.

1:44:46.280,1:44:47.060
OK.

1:44:47.060,1:44:49.730
So I click person.

1:44:49.730,1:44:52.100
Again, if you're not[br]sure, click not sure.

1:44:52.100,1:44:55.100
The point is, just by clicking[br]person and as you can see

1:44:55.100,1:44:57.780
this would work[br]very well on mobile,

1:44:57.780,1:45:01.430
which is why I said you can[br]contribute on your commute.

1:45:01.430,1:45:04.100
You can just hold your[br]phone or tablet or whatever,

1:45:04.100,1:45:05.840
and just tap.

1:45:05.840,1:45:07.040
Person, not a person.

1:45:07.040,1:45:08.900
Person, not a person.

1:45:08.900,1:45:12.500
The amazing thing is that just[br]tapping person has actually

1:45:12.500,1:45:15.830
made an edit to Wikidata[br]on my behalf, which

1:45:15.830,1:45:21.560
I can find out, like every[br]wiki, by clicking contributions.

1:45:21.560,1:45:24.200
And as you can see in addition[br]to the stuff about circus

1:45:24.200,1:45:28.340
Bulgaria, my latest edit is in[br]fact about this Pedro de Ovideo

1:45:28.340,1:45:30.130
Falconi person.

1:45:30.130,1:45:32.000
And the edit was, you can--

1:45:32.000,1:45:38.030
I hope you can see this, created[br]the claim instance of human.

1:45:38.030,1:45:39.110
So I added--

1:45:39.110,1:45:43.100
I mean Wikidata game[br]added for me the statement

1:45:43.100,1:45:44.180
instance of human.

1:45:44.180,1:45:47.780
Now, the awesome thing is[br]that it was super easy to do.

1:45:47.780,1:45:51.890
I didn't have to go into that[br]entity, click the Add button,

1:45:51.890,1:45:57.080
choose the instance of property,[br]choose human, hit Save.

1:45:57.080,1:45:59.210
Instead of all these[br]operations I just

1:45:59.210,1:46:04.250
tapped on my screen,[br]person, not a person.

1:46:04.250,1:46:10.280
And I can do hundreds of[br]edits during my daily commute.

1:46:10.280,1:46:12.410
There are other games,[br]like the gender game.

1:46:12.410,1:46:14.810
So this is about--

1:46:14.810,1:46:17.240
this is when Wikidata[br]already knows

1:46:17.240,1:46:19.760
that this item is a[br]person, but it doesn't

1:46:19.760,1:46:21.710
know the gender of this person.

1:46:21.710,1:46:25.340
Which is another one of[br]the more basic items.

1:46:25.340,1:46:27.770
And this is taking a long[br]time because of the language

1:46:27.770,1:46:29.870
limitations that I set on it.

1:46:29.870,1:46:32.660
I guess the less exotic[br]languages have already

1:46:32.660,1:46:35.130
been exhausted in the game.

1:46:35.130,1:46:36.880
We don't have to[br]wait all this time.

1:46:40.280,1:46:44.970
We can try something else.

1:46:44.970,1:46:45.950
How about occupation?

1:46:45.950,1:46:46.850
The occupation game.

1:46:46.850,1:46:49.400
Here we go, this is in Russian.

1:46:49.400,1:46:55.540
And what is the occupation[br]of this gentleman?

1:46:55.540,1:46:58.630
Well he is an [INAUDIBLE].

1:46:58.630,1:47:00.700
He's a church person.

1:47:00.700,1:47:04.300
However, so the[br]occupation game is

1:47:04.300,1:47:06.490
where Wikidata game[br]will automatically

1:47:06.490,1:47:10.990
pull likely occupations[br]from the article text

1:47:10.990,1:47:13.810
and ask for confirmation.

1:47:13.810,1:47:16.840
So if he-- if this person[br]really is a deacon,

1:47:16.840,1:47:17.770
I should click that.

1:47:17.770,1:47:19.990
But I'm not sure.

1:47:19.990,1:47:24.950
I'm not clear on the Russian[br]church's distinctions between--

1:47:24.950,1:47:26.620
I mean [INAUDIBLE][br]is pretty senior,

1:47:26.620,1:47:28.690
but I don't know if that[br]automatically also means

1:47:28.690,1:47:30.100
he's a deacon or not.

1:47:30.100,1:47:32.720
And [INAUDIBLE] is[br]not listed here.

1:47:32.720,1:47:36.380
So I will click not listed.

1:47:36.380,1:47:39.540
Also, these guesses[br]are not always correct.

1:47:39.540,1:47:42.680
So, this guy for[br]example, is in Russian.

1:47:42.680,1:47:43.430
I can read this.

1:47:43.430,1:47:44.470
He's a philologist.

1:47:44.470,1:47:45.380
He's a linguist.

1:47:45.380,1:47:48.510
So I can confirm it[br]and click linguist.

1:47:48.510,1:47:49.010
All right?

1:47:49.010,1:47:51.950
And again, if we look[br]at my contributions

1:47:51.950,1:47:55.700
we can see the Wikidata[br]game on my behalf

1:47:55.700,1:47:59.930
created occupation linguist.

1:47:59.930,1:48:02.450
OK.

1:48:02.450,1:48:04.370
Just by typing linguist there.

1:48:04.370,1:48:07.040
Now if it's taken[br]from the article,

1:48:07.040,1:48:09.860
why would it ever be wrong?

1:48:09.860,1:48:15.970
Well Jesus was the[br]son of a carpenter.

1:48:15.970,1:48:18.870
The word carpenter[br]appears in the text.

1:48:18.870,1:48:22.840
That doesn't mean it's correct[br]to say Jesus was a carpenter.

1:48:22.840,1:48:23.340
OK?

1:48:23.340,1:48:24.660
Just a trivial example, right?

1:48:24.660,1:48:30.250
So many, many articles will say,[br]you know, born to a physician.

1:48:30.250,1:48:32.850
And so the word physician[br]could be guessed,

1:48:32.850,1:48:36.030
but it wouldn't be correct[br]unless the son is also

1:48:36.030,1:48:38.090
a physician.

1:48:38.090,1:48:43.540
So I hope it gives[br]you the gist of it.

1:48:43.540,1:48:47.500
There is also a[br]distributed Wikidata game,

1:48:47.500,1:48:48.774
which is pretty awesome.

1:48:51.450,1:48:54.320
Here we go, which[br]has additional games.

1:48:54.320,1:49:02.610
So, for example, the[br]key on game gives you,

1:49:02.610,1:49:06.940
maybe it gives you,[br]some items to play with.

1:49:16.610,1:49:17.110
Yes?

1:49:17.110,1:49:17.610
No?

1:49:17.610,1:49:18.430
OK.

1:49:18.430,1:49:20.830
So it gives you[br]this little card,

1:49:20.830,1:49:27.940
and asks you to confirm is this[br]instance of human settlement?

1:49:27.940,1:49:30.480
That is, is it a village,[br]town, city, whatever.

1:49:30.480,1:49:33.310
Is it a kind of human[br]settlement or not?

1:49:33.310,1:49:34.340
Or maybe it's a book.

1:49:34.340,1:49:35.540
Maybe it's a poem.

1:49:35.540,1:49:38.980
Again, so, is it an[br]English settlement?

1:49:38.980,1:49:41.500
And you can click the languages[br]here to see the information.

1:49:41.500,1:49:43.270
So I can click English.

1:49:43.270,1:49:44.572
And indeed the article--

1:49:44.572,1:49:46.030
I mean the actual[br]Wikipedia article

1:49:46.030,1:49:49.360
says Camigji is a[br]town and territory

1:49:49.360,1:49:51.370
in this district in the Congo.

1:49:51.370,1:49:54.640
So yes, this is an instance[br]of human settlement.

1:49:54.640,1:49:57.580
So I clicked yes.

1:49:57.580,1:50:00.460
And just clicking yes[br]again went to that item,

1:50:00.460,1:50:02.740
and added property[br]of human settlement.

1:50:02.740,1:50:05.560
Now the point of[br]all these games is

1:50:05.560,1:50:08.140
these are tools,[br]written by programmers,

1:50:08.140,1:50:12.490
making kind of semi educated[br]guesses about these fairly

1:50:12.490,1:50:14.120
basic properties.

1:50:14.120,1:50:17.770
And they are meant to[br]semi automate, to assist,

1:50:17.770,1:50:23.730
in the accumulation of all[br]these important pieces of data.

1:50:23.730,1:50:26.640
Now every single[br]click here helps

1:50:26.640,1:50:31.000
Wikidata give better[br]results, richer results

1:50:31.000,1:50:32.380
in future queries.

1:50:32.380,1:50:38.130
Again, as of right now[br]Wikidata can include Camigji

1:50:38.130,1:50:42.690
if I ask it, you know, what[br]are some towns in Congo?

1:50:42.690,1:50:44.220
Until now it could not.

1:50:44.220,1:50:46.830
Because it literally[br]didn't know.

1:50:46.830,1:50:51.950
So every time we click male,[br]female, person, not a person,

1:50:51.950,1:50:56.640
make these decisions,[br]we help improve Wikidata

1:50:56.640,1:51:01.560
and enrich the results[br]that we could receive.

1:51:01.560,1:51:04.590
Any questions about this, about[br]kind of micro contributions

1:51:04.590,1:51:07.010
through the Wikidata game?

1:51:07.010,1:51:09.890
If that looks[br]appealing I encourage

1:51:09.890,1:51:12.860
you to go and visit[br]the Wikidata game

1:51:12.860,1:51:15.205
and start contributing[br]in that way.

1:51:19.580,1:51:21.650
There is a question here.

1:51:21.650,1:51:24.650
If I make an article about[br]Circus Bulgaria how should

1:51:24.650,1:51:26.630
I correctly connect them?

1:51:26.630,1:51:28.740
That is an excellent question.

1:51:28.740,1:51:33.090
So once-- so now there is a[br]Wikidata item about that book,

1:51:33.090,1:51:37.650
but there is no Wikipedia[br]article anywhere.

1:51:37.650,1:51:41.460
Now suppose I write one[br]in, Bulgarian maybe,

1:51:41.460,1:51:42.870
you go to Wikidata.

1:51:42.870,1:51:45.180
You find the item by searching.

1:51:45.180,1:51:49.170
You find the item, and then[br]the empty site links section

1:51:49.170,1:51:50.850
right at the bottom there--

1:51:50.850,1:51:52.020
where are we?

1:51:52.020,1:51:53.100
We have this?

1:51:53.100,1:51:55.050
Circus Bulgaria.

1:51:55.050,1:51:56.010
Let's demonstrate this.

1:51:56.010,1:51:58.000
So here is the item[br]about the book.

1:51:58.000,1:52:01.030
Let's say that now[br]there is an article

1:52:01.030,1:52:03.670
because I just created it.

1:52:03.670,1:52:07.450
I can go here to the empty[br]Wikipedia link section,

1:52:07.450,1:52:11.760
click Edit, type the[br]name of the wiki,

1:52:11.760,1:52:16.430
let's say English, and then[br]type the name of the page

1:52:16.430,1:52:18.230
that I just created.

1:52:18.230,1:52:20.790
Circus-- right?

1:52:20.790,1:52:23.400
And again, it offers[br]me auto-complete

1:52:23.400,1:52:25.080
for my convenience.

1:52:25.080,1:52:28.260
Now we don't actually[br]have the article created,

1:52:28.260,1:52:30.480
but I could let's just[br]say this was the article.

1:52:30.480,1:52:33.330
I can just click this,[br]hit Save, and that

1:52:33.330,1:52:36.450
would associate the[br]new Wikipedia article

1:52:36.450,1:52:38.130
with this Wikidata item.

1:52:38.130,1:52:41.940
That is the beginning of the[br]inter-wiki list for this item.

1:52:41.940,1:52:43.620
I will not click[br]Save Now, because we

1:52:43.620,1:52:45.289
didn't have the article yet.

1:52:45.289,1:52:46.830
So I hope that[br]answers that question.

1:52:46.830,1:52:50.340
Was there another question[br]that I missed here?

1:52:50.340,1:52:51.450
No.

1:52:51.450,1:52:53.170
OK.

1:52:53.170,1:52:55.300
Any questions about[br]the Wikidata game?

1:52:55.300,1:53:00.740
About this idea of[br]micro contributions?

1:53:00.740,1:53:05.330
If not then we can move[br]on to embedding data,

1:53:05.330,1:53:07.490
and after that we[br]can discuss queries,

1:53:07.490,1:53:12.000
how to get at all this[br]data from Wikidata.

1:53:12.000,1:53:16.500
So the short version of how[br]to embed data from Wikidata

1:53:16.500,1:53:19.920
is that there is this[br]little magic incantation.

1:53:19.920,1:53:25.410
Curly brace, curly brace,[br]hash mark, property.

1:53:25.410,1:53:29.820
It looks like a template, but[br]it isn't because of that hash.

1:53:29.820,1:53:31.320
And that is magic.

1:53:31.320,1:53:34.170
Take a look at this little[br]demo that I prepared.

1:53:34.170,1:53:37.950
This page, which is off[br]my user page on meta,

1:53:37.950,1:53:40.110
but it could be on any wiki.

1:53:40.110,1:53:42.490
OK.

1:53:42.490,1:53:49.420
Says, since San Francisco[br]is item Q62 in Wikidata,

1:53:49.420,1:53:55.240
and since population is[br]property P1082, I can tell you

1:53:55.240,1:53:58.840
that according to Wikidata the[br]population of San Francisco

1:53:58.840,1:54:02.180
is this.

1:54:02.180,1:54:08.420
And this bolded number here was[br]produced with this incantation.

1:54:08.420,1:54:14.420
Curly brace, curly brace,[br]hash mark, property P1082,

1:54:14.420,1:54:18.751
that's population,[br]type from what item?

1:54:18.751,1:54:19.250
Right?

1:54:19.250,1:54:21.650
Cause I'm pulling[br]an arbitrary number.

1:54:21.650,1:54:23.570
I could put any[br]property in any item

1:54:23.570,1:54:27.020
here, and kind of include[br]it, embedded, into my text.

1:54:27.020,1:54:29.630
This isn't even about-- you[br]notice this is my user page.

1:54:29.630,1:54:32.480
This isn't even the article[br]about San Francisco.

1:54:32.480,1:54:35.210
I just want to pull that[br]number into this thing

1:54:35.210,1:54:36.410
that I'm writing.

1:54:36.410,1:54:38.820
So it's fairly simple.

1:54:38.820,1:54:40.970
I identify the property.

1:54:40.970,1:54:43.440
I identify the item[br]to take it from.

1:54:43.440,1:54:47.120
And Wikidata will,[br]I mean Wikipedia,

1:54:47.120,1:54:50.480
or the wiki I'm on, in this[br]case meta, will go to Wikipedia

1:54:50.480,1:54:52.820
and fetch it for me.

1:54:52.820,1:54:56.480
Likewise, since Denny Vrandecic,[br]the designer of Wikidata

1:54:56.480,1:55:01.370
is item 18618629, right?

1:55:01.370,1:55:04.790
I mean, he's a notable person,[br]so he has a Wikidata entity.

1:55:04.790,1:55:09.160
And since occupation is property[br]106, and date of birth is 569,

1:55:09.160,1:55:12.290
and place of birth[br]is 19, because

1:55:12.290,1:55:14.720
of all that I can tell you[br]that Vrandecic was born

1:55:14.720,1:55:19.130
in Stuttgart, on this date,[br]and is researcher, programmer,

1:55:19.130,1:55:20.850
and computer scientist.

1:55:20.850,1:55:25.010
If you look at the source for[br]this page, click Edit Source,

1:55:25.010,1:55:28.700
you can see that the word[br]Stuttgart does not appear here,

1:55:28.700,1:55:30.530
because it came from Wikidata.

1:55:30.530,1:55:34.171
I did not write this into[br]my little demo page here.

1:55:34.171,1:55:34.670
See?

1:55:34.670,1:55:37.380
Place of birth is--

1:55:37.380,1:55:37.880
where is it?

1:55:37.880,1:55:38.380
Here.

1:55:38.380,1:55:43.790
Born in property 19 from[br]queue number so-and-so.

1:55:43.790,1:55:46.970
That is how easy[br]it is to pull stuff

1:55:46.970,1:55:51.890
into a wiki from Wikidata.

1:55:51.890,1:55:55.280
OK now there's[br]some nuance to it.

1:55:55.280,1:55:57.470
And there's there are[br]some additional parameters

1:55:57.470,1:55:58.130
you can give.

1:55:58.130,1:56:00.230
And you can ask[br]Wikidata to give you

1:56:00.230,1:56:03.635
not just the text of the values,[br]but actually make it links.

1:56:06.750,1:56:14.825
So, for example, if I change[br]this from property to values--

1:56:25.950,1:56:29.142
No, that did not work at all.

1:56:29.142,1:56:29.850
Wasn't it values?

1:56:29.850,1:56:30.350
What was it?

1:56:33.370,1:56:34.614
Values and then--

1:57:19.265,1:57:19.890
Oh, statements.

1:57:19.890,1:57:20.710
My bad, sorry.

1:57:20.710,1:57:22.980
The Magic word is statements.

1:57:22.980,1:57:24.010
Statements.

1:57:24.010,1:57:28.680
So going back here.

1:57:28.680,1:57:35.385
If I change the word property[br]to the word statements

1:57:35.385,1:57:40.890
here then this same value--

1:57:40.890,1:57:43.300
that did not work at all.

1:57:43.300,1:57:46.690
Oh, because I'm on meta.

1:57:46.690,1:57:48.670
So because I'm on[br]meta, meta doesn't

1:57:48.670,1:57:52.230
have an article named[br]researcher, programmer,

1:57:52.230,1:57:53.500
or computer scientist.

1:57:53.500,1:57:55.120
But Wikipedia does.

1:57:55.120,1:58:00.210
If I included this same[br]syntax in Wikipedia,

1:58:00.210,1:58:02.950
like English Wikipedia,[br]for example--

1:58:02.950,1:58:04.855
So let's go there right now.

1:58:11.240,1:58:13.480
And go-- go to my--

1:58:18.550,1:58:19.345
Go to my sandbox.

1:58:23.090,1:58:27.982
If I just brutally paste[br]this on my sandbox here--

1:58:32.690,1:58:35.810
So, see, these became links.

1:58:35.810,1:58:39.740
Because Wikipedia has an article[br]called programmer and computer

1:58:39.740,1:58:40.910
scientist.

1:58:40.910,1:58:43.460
So, like I said, there's[br]some additional nuance

1:58:43.460,1:58:44.840
to the embedding.

1:58:44.840,1:58:47.030
The important thing[br]is that this is

1:58:47.030,1:58:51.470
the key to delivering on that[br]first problem that I mentioned.

1:58:51.470,1:58:55.970
How to get data from[br]a central location

1:58:55.970,1:58:58.850
onto your wiki in your language.

1:58:58.850,1:59:04.460
Basically using property and[br]statements magic incantations.

1:59:04.460,1:59:07.100
And of course,[br]usually, this would be

1:59:07.100,1:59:10.010
in the context of an info box.

1:59:10.010,1:59:14.180
Some wikis-- English Wikipedia[br]is not leading the way there.

1:59:14.180,1:59:16.490
Some smaller wikis[br]are more advanced

1:59:16.490,1:59:22.070
actually in integrating[br]Wikidata embeddings like this

1:59:22.070,1:59:24.620
into their info boxes.

1:59:24.620,1:59:26.300
So that instead of[br]the info box just

1:59:26.300,1:59:30.620
being a template on the wiki[br]with field equals value,

1:59:30.620,1:59:31.685
field equals value.

1:59:31.685,1:59:35.700
That template of the[br]info box on the wiki

1:59:35.700,1:59:40.160
pulls the values, the birthdate,[br]the languages, et cetera,

1:59:40.160,1:59:44.210
pulls them from Wikidata.

1:59:44.210,1:59:49.820
So basically just-- I just[br]demonstrated single calls

1:59:49.820,1:59:52.550
to this, but of course[br]an info box template

1:59:52.550,1:59:56.270
would include maybe[br]20 or 40 such embeds,

1:59:56.270,1:59:57.710
and that is not a problem.

1:59:57.710,2:00:01.460
Of course, before you go and[br]edit the English Wikipedia's

2:00:01.460,2:00:06.050
info box person and replace[br]it all with Wikidata embeds,

2:00:06.050,2:00:09.050
you should discuss it with the[br]English Wikipedia community.

2:00:09.050,2:00:12.000
These discussions have[br]already been taking place.

2:00:12.000,2:00:13.640
There are some[br]concerns about how

2:00:13.640,2:00:17.150
to patrol this, how to keep[br]it newbie friendly, et cetera.

2:00:17.150,2:00:20.690
So there are legitimate concerns[br]with just moving everything

2:00:20.690,2:00:22.910
to be embedded from Wikidata.

2:00:22.910,2:00:26.450
But the communities are[br]gradually handling this.

2:00:26.450,2:00:29.390
I mean this ability to embed[br]from Wikidata is not very old.

2:00:29.390,2:00:31.550
It's been around[br]for about a year.

2:00:31.550,2:00:35.150
So communities are[br]still working on kind

2:00:35.150,2:00:37.560
of integrating that technology.

2:00:37.560,2:00:40.190
But that is that is kind[br]of just the basics of how

2:00:40.190,2:00:44.210
to pull data, individual bits[br]of data, that's not querying,

2:00:44.210,2:00:47.330
that's not asking those sweeping[br]questions that I was talking

2:00:47.330,2:00:48.850
about yet.

2:00:48.850,2:00:50.720
We'll get to that[br]right now this is

2:00:50.720,2:00:55.310
how to pull a specific datum,[br]a specific piece of data,

2:00:55.310,2:00:57.395
from Wikidata.

2:01:01.530,2:01:02.530
OK.

2:01:02.530,2:01:07.080
So here's another quick[br]thing to demonstrate

2:01:07.080,2:01:09.880
before we go to[br]queries, and that

2:01:09.880,2:01:12.010
is the article placeholder.

2:01:12.010,2:01:15.010
The article placeholder[br]is a feature

2:01:15.010,2:01:19.660
that is being tested on the[br]Esperanto Wikipedia, and maybe

2:01:19.660,2:01:22.180
another wiki, I don't remember.

2:01:22.180,2:01:28.490
And it is using the[br]potential of Wikidata

2:01:28.490,2:01:32.690
to offer a placeholder[br]for an article.

2:01:32.690,2:01:37.940
An automatically generated[br]Wikidata powered replacement

2:01:37.940,2:01:41.720
placeholder for an article[br]for articles that don't yet

2:01:41.720,2:01:45.950
exist on Esperanto.

2:01:45.950,2:01:50.440
So let's go to the[br]Esperanto Wikipedia.

2:01:50.440,2:01:52.440
I don't speak Esperanto.

2:01:52.440,2:01:56.760
But let's look for Helen[br]Dewitt, our friend,

2:01:56.760,2:01:58.170
in Esperanto Wikipedia.

2:01:58.170,2:02:00.270
Now Esperanto is not[br]one of the Wikipedias

2:02:00.270,2:02:03.060
that have an article[br]about Helen Dewitt.

2:02:03.060,2:02:04.890
And so it tells me that, right?

2:02:04.890,2:02:06.570
There is no Helen Dewitt.

2:02:06.570,2:02:08.670
Maybe you were looking[br]for Helena Dewitt.

2:02:08.670,2:02:10.200
No, I was not.

2:02:10.200,2:02:13.650
You can start an article[br]about Helen Dewitt.

2:02:13.650,2:02:15.390
You can search.

2:02:15.390,2:02:17.820
You know, there's[br]all this stuff.

2:02:17.820,2:02:24.180
But there is also this[br]little option here, hiding,

2:02:24.180,2:02:30.640
which tells me that the[br]Esperanto Wikipedia is--

2:02:30.640,2:02:31.580
what's happening here?

2:02:35.140,2:02:35.890
Yes.

2:02:35.890,2:02:40.520
The Esperanto Wikipedia is[br]ready to give me this page.

2:02:40.520,2:02:44.020
This page, as you can see, it's[br]on the Esperanto Wikipedia,

2:02:44.020,2:02:46.090
but it's not an article.

2:02:46.090,2:02:47.480
See, it's a special page.

2:02:47.480,2:02:49.700
It's machine generated.

2:02:49.700,2:02:52.150
You can see the URL as well.

2:02:52.150,2:02:54.410
It's not, you know,[br]slash Helen Dewitt.

2:02:54.410,2:02:58.450
It's slash specialio,[br]about topic,

2:02:58.450,2:03:01.570
and then the Wikidata[br]ID of Helen Dewitt.

2:03:01.570,2:03:03.760
And what I get here--

2:03:03.760,2:03:05.860
I get an English[br]description, by the way,

2:03:05.860,2:03:08.300
because there is no[br]Esperanto description.

2:03:08.300,2:03:10.420
Wikidata can't make it up.

2:03:10.420,2:03:13.600
But what it can do is[br]offer me these pieces

2:03:13.600,2:03:16.960
of data in my language,[br]in this case Esperanto.

2:03:16.960,2:03:18.921
I'm on the Esperanto Wikipedia.

2:03:18.921,2:03:19.420
OK.

2:03:19.420,2:03:23.380
So it tells me that she's[br]American, for example,

2:03:23.380,2:03:26.090
and it tells me[br]that in Esperanto.

2:03:26.090,2:03:29.350
OK and it tells me[br]that she speaks Latin.

2:03:29.350,2:03:32.410
Remember we taught[br]Wikidata that?

2:03:32.410,2:03:35.800
It tells me that she[br]was educated in Oxford,

2:03:35.800,2:03:38.050
you know, and gives me the[br]references to the extent

2:03:38.050,2:03:39.130
that they exist.

2:03:39.130,2:03:41.560
I mean this is not an article.

2:03:41.560,2:03:46.650
It's not, you know, paragraphs[br]of fluent Esperanto text.

2:03:46.650,2:03:50.190
But it is information[br]that I can understand

2:03:50.190,2:03:51.960
if I speak this language.

2:03:51.960,2:03:55.380
And it's better than nothing.

2:03:55.380,2:04:00.120
And remember Helen Dewitt was[br]not a very detailed article.

2:04:00.120,2:04:03.690
If I were to ask about, I[br]don't know, some politician,

2:04:03.690,2:04:08.340
or popular singer that[br]has more data in Wikidata,

2:04:08.340,2:04:12.690
than this machine generated[br]thing would have been richer.

2:04:12.690,2:04:16.320
So this feature is available[br]and is under beta testing

2:04:16.320,2:04:19.530
right now, but generally if[br]this sounds interesting for you

2:04:19.530,2:04:21.600
especially if you come[br]from a smaller wiki that

2:04:21.600,2:04:25.230
is missing a lot of articles[br]that people may want to learn

2:04:25.230,2:04:28.320
about, you can contact[br]the Wikimedia foundation

2:04:28.320,2:04:33.486
and ask for article placeholder[br]to be enabled on your wiki.

2:04:33.486,2:04:34.860
And again, this[br]is a placeholder.

2:04:34.860,2:04:37.890
Of course, it exists only[br]until someone actually

2:04:37.890,2:04:43.290
writes a proper Esperanto[br]article about Helen Dewitt.

2:04:43.290,2:04:45.060
So I hope this is clear.

2:04:45.060,2:04:50.810
This is all coming from[br]Wikidata on the fly.

2:04:50.810,2:04:51.470
In real time.

2:04:51.470,2:04:57.500
As you can see it includes my[br]latest edits to Helen Dewitt.

2:04:57.500,2:04:58.940
OK.

2:04:58.940,2:05:05.250
Questions about the-- questions[br]about the article placeholder?

2:05:05.250,2:05:09.580
If there are try and[br]put them on the channel.

2:05:09.580,2:05:13.300
And this brings us to one of[br]the main courses of this talk,

2:05:13.300,2:05:15.270
which is querying Wikidata.

2:05:15.270,2:05:18.660
So I've explained[br]how Wikidata works.

2:05:18.660,2:05:19.680
We've walked through it.

2:05:19.680,2:05:20.850
We've added to it.

2:05:20.850,2:05:22.800
We've created a new item.

2:05:22.800,2:05:26.360
We learned how to contribute[br]during our commutes.

2:05:26.360,2:05:30.150
And all this was you[br]kept promising us,

2:05:30.150,2:05:32.050
Asaf, that this would be--

2:05:32.050,2:05:34.690
this would enable[br]these amazing queries.

2:05:34.690,2:05:37.960
So time to make good on that.

2:05:37.960,2:05:42.880
The URL you need to remember[br]is query.wikidata.org.

2:05:42.880,2:05:49.390
And that will take you[br]to a query system that

2:05:49.390,2:05:52.510
uses a language called SPARQL.

2:05:52.510,2:05:58.150
SPARQL, spelt with[br]a Q. This language

2:05:58.150,2:06:01.690
is not a Wikimedia creation.

2:06:01.690,2:06:06.010
It's a standardized language[br]used for querying linked data

2:06:06.010,2:06:07.540
sources.

2:06:07.540,2:06:10.720
And because of that[br]there are there

2:06:10.720,2:06:14.590
are certain usability prices[br]that we pay for using SPARQL,

2:06:14.590,2:06:16.010
for using a standard language.

2:06:16.010,2:06:19.570
It's not completely custom[br]made for querying Wikidata,

2:06:19.570,2:06:21.740
and we'll see that[br]in just a moment.

2:06:21.740,2:06:23.530
The principle to[br]remember about Wikidata

2:06:23.530,2:06:27.880
query is that Wikidata will[br]tell you everything it knows,

2:06:27.880,2:06:29.470
but no more.

2:06:29.470,2:06:32.440
I have anticipated this[br]several times already, right?

2:06:32.440,2:06:35.980
Until this moment when[br]we taught Wikidata data

2:06:35.980,2:06:38.590
that Helen Dewitt[br]speaks Latin, she

2:06:38.590,2:06:41.500
would not have appeared[br]in query results

2:06:41.500,2:06:45.974
asking who are American[br]writers who speak Latin?

2:06:45.974,2:06:47.140
She would not have appeared.

2:06:47.140,2:06:49.090
But as of this[br]afternoon, she will

2:06:49.090,2:06:52.950
appear because I've added[br]that piece of information.

2:06:52.950,2:07:01.380
So a result of that principle[br]is that you can never say,

2:07:01.380,2:07:05.950
well I ran a Wikidata[br]query and this

2:07:05.950,2:07:11.510
is the list of Flemish painters[br]who are sons of painters.

2:07:11.510,2:07:12.310
The list.

2:07:12.310,2:07:14.110
That these are all[br]the Flemish painters

2:07:14.110,2:07:15.220
who are sons of painters.

2:07:15.220,2:07:19.390
That is never something you can[br]say based on a Wikidata query,

2:07:19.390,2:07:22.390
because of course, maybe[br]not all the Flemish painters

2:07:22.390,2:07:26.020
who are sons of painters have[br]been expressed in Wikidata data

2:07:26.020,2:07:26.760
yet.

2:07:26.760,2:07:28.840
Wikidata doesn't know[br]about some of them,

2:07:28.840,2:07:30.340
or maybe it knows[br]about all of them

2:07:30.340,2:07:32.500
but doesn't know[br]the important fact

2:07:32.500,2:07:35.200
that this person is[br]the son of that person,

2:07:35.200,2:07:38.740
because those properties[br]have not been added.

2:07:38.740,2:07:40.940
And so they cannot be[br]included in the results.

2:07:40.940,2:07:42.550
So the results of[br]a Wikidata query

2:07:42.550,2:07:46.870
are never the definitive sets.

2:07:46.870,2:07:49.600
What you can say about[br]a Wikidata query is here

2:07:49.600,2:07:52.840
are some Flemish painters[br]who are sons of painters.

2:07:52.840,2:07:56.260
Here are some cities[br]with female mayors.

2:07:56.260,2:07:58.270
Whatever it is[br]you're querying about

2:07:58.270,2:08:01.030
is never guaranteed[br]to be complete

2:08:01.030,2:08:03.580
because Wikidata,[br]like Wikipedia, is

2:08:03.580,2:08:05.530
a work in progress.

2:08:05.530,2:08:13.240
And of course, the more[br]we teach Wikidata the

2:08:13.240,2:08:16.240
more useful it becomes.

2:08:16.240,2:08:22.520
OK so lets go and[br]see those queries.

2:08:22.520,2:08:25.990
So this is query.wikidata.org.

2:08:25.990,2:08:29.000
It's not the wiki.

2:08:29.000,2:08:29.500
All right?

2:08:29.500,2:08:32.530
So this isn't like some[br]page on the wiki itself.

2:08:32.530,2:08:35.099
This is kind of an[br]external system.

2:08:35.099,2:08:35.890
So it's not a wiki.

2:08:35.890,2:08:37.960
You can see I don't[br]have a user page here.

2:08:37.960,2:08:39.520
I don't have a history tab.

2:08:39.520,2:08:40.960
This isn't a wiki page.

2:08:40.960,2:08:44.560
This is a special kind[br]of tool or system.

2:08:44.560,2:08:51.330
And it invites me to[br]input a SPARQL query.

2:08:51.330,2:08:55.060
Now most of us do[br]not speak SPARQL.

2:08:55.060,2:08:59.800
It's a a technical language.

2:08:59.800,2:09:01.720
It's a query language.

2:09:01.720,2:09:06.760
Some of you may be thinking[br]about SQL, the database query

2:09:06.760,2:09:08.500
language.

2:09:08.500,2:09:13.330
SPARQL is named with kind[br]of a wink, or a nod, to SQL.

2:09:13.330,2:09:17.440
But, I warn you, if[br]you are comfortable in

2:09:17.440,2:09:22.750
SQL don't expect to carry[br]over your knowledge of SQL

2:09:22.750,2:09:23.550
into SPARQL.

2:09:23.550,2:09:26.140
They're not the same.

2:09:26.140,2:09:27.940
They are superficially similar.

2:09:27.940,2:09:28.440
Right?

2:09:28.440,2:09:31.530
So they both use[br]the keyword select,

2:09:31.530,2:09:35.010
and they use the word where,[br]and they use things like limit,

2:09:35.010,2:09:35.770
and order.

2:09:35.770,2:09:38.190
So again, if you know[br]this already from SQL

2:09:38.190,2:09:40.500
those mean roughly[br]the same things,

2:09:40.500,2:09:44.550
but don't expect it to[br]behave just like SQL.

2:09:44.550,2:09:49.800
You do need to spend some time[br]understanding how SPARQL works.

2:09:49.800,2:09:52.560
So, by all means, I[br]invite you to go and read

2:09:52.560,2:09:55.680
one of the many fine[br]SPARQL tutorials that

2:09:55.680,2:09:59.590
are out there on the web, or[br]to click the Help button here,

2:09:59.590,2:10:03.930
which also includes[br]help about SPARQL.

2:10:03.930,2:10:08.440
But I also know[br]that most of us when

2:10:08.440,2:10:12.580
we want to do some advanced[br]formatting on wiki,

2:10:12.580,2:10:16.090
for example, we don't go[br]and read the help page

2:10:16.090,2:10:18.220
on templates, right?

2:10:18.220,2:10:21.460
We go to a page that already[br]does what we want to do,

2:10:21.460,2:10:27.430
and adopt and adapt the code[br]from that other page, right?

2:10:27.430,2:10:30.610
So we just take something that[br]does roughly what we want,

2:10:30.610,2:10:33.280
and just copy it over and[br]change what we need to change.

2:10:33.280,2:10:35.620
That is a very pragmatic[br]and reasonable way

2:10:35.620,2:10:37.420
to do things which is why--

2:10:37.420,2:10:39.850
and the wiki data[br]engineers know this,

2:10:39.850,2:10:43.300
which is why they prepared[br]this very handy button for us

2:10:43.300,2:10:45.580
called examples.

2:10:45.580,2:10:47.710
We click the examples button.

2:10:47.710,2:10:52.390
And, oh my god, there is a ton[br]of-- well there's 312 example

2:10:52.390,2:10:55.582
queries for us to choose from.

2:10:55.582,2:10:57.040
And we can just[br]pick something that

2:10:57.040,2:11:00.310
is roughly like what[br]we're trying to find out,

2:11:00.310,2:11:02.740
and then just change[br]what needs changing.

2:11:02.740,2:11:05.410
So let's take a very simple one.

2:11:05.410,2:11:07.020
The cats query.

2:11:07.020,2:11:10.270
Maybe one of the simplest[br]you could possibly have.

2:11:10.270,2:11:13.510
And let's run it first[br]and then I'll kind of

2:11:13.510,2:11:16.420
walk you through it.

2:11:16.420,2:11:18.460
The goal here is not[br]to teach you SPARQL,

2:11:18.460,2:11:20.860
but to get you to be kind[br]of literate in SPARQL.

2:11:20.860,2:11:23.980
To kind of understand why[br]this does what it does.

2:11:23.980,2:11:25.730
So let's run this query first.

2:11:25.730,2:11:31.390
We click Run and here I[br]have results at the bottom.

2:11:31.390,2:11:34.060
The item, which is[br]just a Wikidata item,

2:11:34.060,2:11:35.290
which of course is a number.

2:11:35.290,2:11:38.860
Remember, wiki data thinks[br]of items as queue numbers.

2:11:38.860,2:11:40.900
And the label,[br]because we're humans

2:11:40.900,2:11:43.190
and we prefer words to numbers.

2:11:43.190,2:11:49.870
So these 114 results[br]are all the cats

2:11:49.870,2:11:53.310
that wiki data knows about.

2:11:53.310,2:11:55.380
Is this all the[br]cats in the world?

2:11:55.380,2:11:57.320
No of course not, remember?

2:11:57.320,2:11:59.730
It's all the cats Wikidata[br]knows about, which

2:11:59.730,2:12:01.410
means they're somehow notable.

2:12:01.410,2:12:05.130
I mean someone bothered to[br]describe them on Wikidata.

2:12:05.130,2:12:12.570
And Wikidata was told this[br]item is an instance of cat.

2:12:12.570,2:12:13.620
Right?

2:12:13.620,2:12:17.040
So these are those cats.

2:12:17.040,2:12:18.540
And we can click any of them.

2:12:18.540,2:12:20.190
I don't know,[br]Pixel, for example.

2:12:20.190,2:12:21.780
Click the Wikipedia item.

2:12:21.780,2:12:24.090
And here is the Wikidata[br]item about Pixel

2:12:24.090,2:12:25.860
with the queue number.

2:12:25.860,2:12:28.980
And he is a tortoiseshell cat.

2:12:28.980,2:12:32.640
And as you can see[br]instance of cat.

2:12:32.640,2:12:33.610
OK.

2:12:33.610,2:12:37.220
And he is five inches high.

2:12:37.220,2:12:41.780
And he is apparently documented[br]in Indonesian, In Bahasa.

2:12:41.780,2:12:45.080
Right here this is Pixel.

2:12:45.080,2:12:50.060
And he is apparently somehow[br]related to the Guinness World

2:12:50.060,2:12:52.160
Records book.

2:12:52.160,2:12:54.650
I don't speak Bahasa, so[br]I don't know exactly why

2:12:54.650,2:12:56.120
this cat is so notable.

2:12:56.120,2:12:58.889
But, of course, cats[br]can become notable

2:12:58.889,2:12:59.930
for all kinds of reasons.

2:12:59.930,2:13:02.204
Maybe they're a[br]YouTube sensation,

2:13:02.204,2:13:03.620
you know, maybe[br]they were involved

2:13:03.620,2:13:05.330
in some historical event.

2:13:05.330,2:13:09.410
I like this cat named Gladstone.

2:13:09.410,2:13:16.590
This cat named Gladstone is--

2:13:16.590,2:13:19.950
he has position[br]held Chief Mouser

2:13:19.950,2:13:22.320
to Her Majesty's Treasury.

2:13:22.320,2:13:25.230
This is an official[br]cat with a job.

2:13:25.230,2:13:29.190
And he has been holding this[br]job, mind you, since the 28th

2:13:29.190,2:13:31.570
of June this past year.

2:13:31.570,2:13:32.970
That's the start time.

2:13:32.970,2:13:35.760
And there is no end time[br]which means he currently

2:13:35.760,2:13:38.850
holds the position[br]of Chief Mouser

2:13:38.850,2:13:40.470
to her Majesty's Treasury.

2:13:40.470,2:13:42.750
His employer is Her[br]Majesty's Treasury.

2:13:42.750,2:13:44.290
He's a male creature.

2:13:44.290,2:13:46.650
And Wikidata knows[br]that this cat is

2:13:46.650,2:13:53.127
named after William Gladstone,[br]the Victorian prime minister.

2:13:53.127,2:13:54.960
Of course if I don't[br]know who this person is

2:13:54.960,2:13:57.540
I can click through[br]and learn that he

2:13:57.540,2:14:01.860
was a liberal politician[br]and prime minister, right?

2:14:01.860,2:14:03.390
He even has a Twitter account.

2:14:03.390,2:14:05.910
And Wikidata sends[br]me right to it.

2:14:05.910,2:14:08.040
The treasury cat[br]Twitter account.

2:14:08.040,2:14:11.010
And he has articles in[br]German, and English,

2:14:11.010,2:14:15.520
and of course Japanese,[br]because he's a cat.

2:14:15.520,2:14:16.020
All right.

2:14:16.020,2:14:19.500
So this was a very simple query.

2:14:19.500,2:14:21.400
Let's find out why it works.

2:14:21.400,2:14:21.900
OK.

2:14:21.900,2:14:25.800
So what did we actually[br]tell Wikidata to do for us?

2:14:25.800,2:14:31.650
We said, please select[br]some items for us

2:14:31.650,2:14:33.580
along with their labels.

2:14:33.580,2:14:34.080
OK?

2:14:34.080,2:14:36.180
Along with their[br]human readable labels

2:14:36.180,2:14:42.010
because if I remove this[br]label what I get is, see,

2:14:42.010,2:14:44.200
just a list of item numbers.

2:14:44.200,2:14:45.280
That's not as fun.

2:14:45.280,2:14:46.930
So that's what this[br]little bit did.

2:14:46.930,2:14:49.630
I just said, give me the[br]items, but also they're

2:14:49.630,2:14:52.330
human readable label.

2:14:52.330,2:14:54.620
And I want you to[br]select a bunch of items,

2:14:54.620,2:14:56.770
but not just any[br]random bunch of items,

2:14:56.770,2:15:01.210
I want to select items where[br]a certain condition holds.

2:15:01.210,2:15:02.790
What is the condition?

2:15:02.790,2:15:06.430
The condition is that the[br]item that I want you to select

2:15:06.430,2:15:14.360
needs to have property[br]31 with a value of Q146.

2:15:14.360,2:15:15.670
Well, that's helpful.

2:15:15.670,2:15:18.070
If I hover over these numbers--

2:15:18.070,2:15:19.750
Again, I get the human[br]readable version.

2:15:19.750,2:15:23.530
So I'm looking for[br]items that have property

2:15:23.530,2:15:28.841
instance of with the value cat.

2:15:28.841,2:15:29.340
Right?

2:15:29.340,2:15:31.173
Because that's literally[br]what I want, right?

2:15:31.173,2:15:33.960
I want all the items that have[br]a property, a statement, that

2:15:33.960,2:15:36.840
says instance of cat.

2:15:36.840,2:15:37.950
That's the condition.

2:15:37.950,2:15:41.640
I'm not interested in items[br]that are instance of book,

2:15:41.640,2:15:43.200
or instance of human.

2:15:43.200,2:15:46.290
I'm interested in[br]instance of cat.

2:15:46.290,2:15:51.090
That is the only condition[br]here in this query.

2:15:51.090,2:15:55.800
This complicated line I ask[br]you to basically ignore.

2:15:55.800,2:15:57.510
This is one of those[br]sacrifices that we

2:15:57.510,2:16:00.720
make for using a standard[br]language like SPARQL.

2:16:00.720,2:16:02.820
But the role of this[br]complicated line

2:16:02.820,2:16:04.920
is to basically[br]ensure that we get

2:16:04.920,2:16:07.860
the English label for that cat.

2:16:07.860,2:16:08.817
OK?

2:16:08.817,2:16:09.900
So don't worry about that.

2:16:09.900,2:16:11.550
Just leave it there.

2:16:11.550,2:16:13.320
And we run the query[br]and we get the list

2:16:13.320,2:16:17.330
of cats with their English[br]labels, and that is awesome.

2:16:17.330,2:16:21.510
By the way, if I change EN,[br]without really understanding

2:16:21.510,2:16:27.260
this line, if I change[br]EN to HE, for Hebrew,

2:16:27.260,2:16:30.160
I get the same results[br]with a Hebrew label.

2:16:30.160,2:16:33.670
Of course, these cats,[br]nobody bothered to give them

2:16:33.670,2:16:35.709
Hebrew labels unfortunately.

2:16:35.709,2:16:37.570
So I get the queue number.

2:16:37.570,2:16:42.874
But if I changed[br]it to Japanese, JA,

2:16:42.874,2:16:45.290
I would get still a bunch of[br]queue numbers for where there

2:16:45.290,2:16:47.389
isn't a Japanese label,[br]but I would get the labels

2:16:47.389,2:16:48.781
in Japanese.

2:16:48.781,2:16:49.280
OK?

2:16:49.280,2:16:51.260
So this is an example[br]of how you don't even

2:16:51.260,2:16:54.620
need to understand all[br]the syntax of this query

2:16:54.620,2:16:56.100
to adapt it to your needs.

2:16:56.100,2:16:58.070
If you want this[br]query as is, but you

2:16:58.070,2:17:00.320
want the labels in[br]Japanese, you can just

2:17:00.320,2:17:03.190
change the language code here.

2:17:03.190,2:17:06.559
OK so that is all[br]this query does.

2:17:06.559,2:17:08.870
Again, just give[br]me the items that

2:17:08.870,2:17:17.590
have property 31, instance of,[br]with a value 146, which is cat.

2:17:17.590,2:17:20.379
Let's take a question just[br]about this very simple query

2:17:20.379,2:17:25.809
before we advance to[br]more complicated queries.

2:17:25.809,2:17:29.200
Any questions just about this?

2:17:29.200,2:17:32.850
Like, did anyone kind of[br]really lose me talking

2:17:32.850,2:17:35.010
about this simple query?

2:17:35.010,2:17:39.389
Again, this query just tells[br]Wikidata, get me all the items

2:17:39.389,2:17:41.280
that somewhere among[br]their statements

2:17:41.280,2:17:44.219
have instance of cat.

2:17:44.219,2:17:46.670
That's the only condition.

2:17:46.670,2:17:47.740
No questions.

2:17:47.740,2:17:49.959
OK, feel free to ask if[br]you'd come up with one.

2:17:49.959,2:17:54.709
So let's complicate[br]things a little.

2:17:54.709,2:17:59.365
Let's ask only for male cats.

2:18:02.080,2:18:03.070
OK.

2:18:03.070,2:18:07.330
Remember this cat[br]Gladstone is male,

2:18:07.330,2:18:09.850
and we know this because[br]he has a property called

2:18:09.850,2:18:14.320
sex or gender, and the value[br]is male creature, right?

2:18:14.320,2:18:17.950
So let's add another[br]condition right here

2:18:17.950,2:18:19.860
under the first condition.

2:18:19.860,2:18:20.870
OK?

2:18:20.870,2:18:22.750
This is a new line.

2:18:22.750,2:18:24.940
And I'm adding a new[br]condition to the query.

2:18:24.940,2:18:30.520
I'm saying, not only do I[br]want this item that you return

2:18:30.520,2:18:35.469
to be instance of cat, I[br]also want this same item

2:18:35.469,2:18:39.280
to have another property,[br]the property sex or gender.

2:18:39.280,2:18:40.299
Right?

2:18:40.299,2:18:43.480
And I need to refer to[br]the property by number.

2:18:43.480,2:18:45.760
But don't worry,[br]Wikidata will help you.

2:18:45.760,2:18:49.500
So you start with this[br]prefix, Wikidata WDDT.

2:18:52.520,2:18:54.980
Again, just ignore[br]that prefix it's

2:18:54.980,2:18:58.940
one of the features of SPARQL[br]that we need to respect.

2:18:58.940,2:19:02.715
WDT colon, and then I can[br]just type control space

2:19:02.715,2:19:04.340
to do a search, to[br]do an auto complete.

2:19:04.340,2:19:08.090
So I can just type sex[br]and Wikidata helpfully

2:19:08.090,2:19:11.760
offers me a drop down[br]with relevant properties.

2:19:11.760,2:19:15.200
So I click property 21, which[br]is the sex or gender property.

2:19:15.200,2:19:17.629
And then I say, so I want[br]the sex or gender property

2:19:17.629,2:19:19.670
to have the Wikidata value.

2:19:19.670,2:19:21.799
Again, control space.

2:19:21.799,2:19:25.340
And I can just[br]say male creature.

2:19:25.340,2:19:25.850
See?

2:19:25.850,2:19:30.950
There's a different item[br]for male, as inhuman,

2:19:30.950,2:19:33.799
and a different one for[br]male creature, for reasons

2:19:33.799,2:19:34.910
that we won't go into.

2:19:34.910,2:19:36.535
Let's pick male[br]creature, because we're

2:19:36.535,2:19:38.040
talking about cats here.

2:19:38.040,2:19:38.540
All right.

2:19:38.540,2:19:42.080
And add a period here at[br]the end and click Run.

2:19:42.080,2:19:48.330
And instead of 114 cats, we get,[br]this time, we got 43 results.

2:19:48.330,2:19:53.360
Including our friend Gladstone[br]who is a male creature cat.

2:19:53.360,2:19:58.530
So that means all the[br]rest are female, right?

2:19:58.530,2:20:00.410
Wrong.

2:20:00.410,2:20:00.980
Wrong.

2:20:00.980,2:20:02.840
That does not mean that at all.

2:20:02.840,2:20:06.530
What it means is of[br]the 114 items that

2:20:06.530,2:20:11.960
have instance of cat,[br]only 43 have explicitly

2:20:11.960,2:20:14.690
sex male creature.

2:20:14.690,2:20:17.570
The rest of them do not.

2:20:17.570,2:20:21.800
Maybe because they have[br]sex female creature,

2:20:21.800,2:20:25.930
but maybe because they don't[br]have that property at all.

2:20:25.930,2:20:28.290
I'm emphasizing[br]this to kind of help

2:20:28.290,2:20:31.770
you train yourself to[br]correctly interpret

2:20:31.770,2:20:34.140
the results of[br]queries from Wikidata.

2:20:34.140,2:20:36.870
Don't jump into this kind[br]of simplistic conclusion,

2:20:36.870,2:20:41.820
OK there's 114 total, 43 male,[br]therefore the rest are female.

2:20:41.820,2:20:43.520
That is not correct.

2:20:43.520,2:20:45.030
OK?

2:20:45.030,2:20:49.740
But 43 of those explicitly[br]had another statement, sex

2:20:49.740,2:20:52.530
or gender, male creature.

2:20:52.530,2:20:55.020
So I just added[br]another condition,

2:20:55.020,2:20:58.290
and now my query is[br]asking two separate things

2:20:58.290,2:21:00.150
about the results.

2:21:00.150,2:21:04.472
They need to be a cat[br]and a male creature.

2:21:04.472,2:21:06.270
AUDIENCE: Maybe we[br]should see how many

2:21:06.270,2:21:08.100
cats have Twitter accounts.

2:21:08.100,2:21:11.440
But there is a[br]question from YouTube,

2:21:11.440,2:21:14.220
which is will you talk about[br]the export possibilities

2:21:14.220,2:21:17.280
of the result of the query?

2:21:17.280,2:21:18.420
ASAF BARTOV: Absolutely.

2:21:18.420,2:21:21.000
Absolutely I will in[br]just a little bit.

2:21:21.000,2:21:23.010
I mean there is, in[br]addition to just getting

2:21:23.010,2:21:28.350
this kind of table, I can get[br]these results in other formats.

2:21:28.350,2:21:30.360
And I can also[br]download these results.

2:21:30.360,2:21:32.820
I can click the Download[br]button and get them

2:21:32.820,2:21:35.070
as a comma separated[br]file, tab separated

2:21:35.070,2:21:38.910
file, a JSON file, which is[br]useful for programmatic uses.

2:21:38.910,2:21:40.590
I can also get a link.

2:21:40.590,2:21:42.330
So I can get a[br]link to this query.

2:21:42.330,2:21:45.990
I mean, I spent all this time[br]designing this beautiful query.

2:21:45.990,2:21:50.280
I can get a short URL that was[br]generated especially for me

2:21:50.280,2:21:52.170
right now with a tiny URL.

2:21:52.170,2:21:54.690
I can just paste this[br]into Twitter and go,

2:21:54.690,2:21:59.280
hey people look at all the male[br]cats that Wikidata knows about.

2:21:59.280,2:22:01.170
OK, this is not a[br]very exciting query.

2:22:01.170,2:22:03.900
But once I get to a really[br]complicated exciting query

2:22:03.900,2:22:07.650
I can totally share that[br]very easily through this.

2:22:07.650,2:22:09.750
And we will get to more[br]interesting queries

2:22:09.750,2:22:11.740
in just a second.

2:22:11.740,2:22:16.400
Any questions on this kind[br]of basic querying so far?

2:22:16.400,2:22:17.940
OK.

2:22:17.940,2:22:25.340
So that was a very[br]simple example.

2:22:25.340,2:22:30.250
Let's spend a moment exploring.

2:22:30.250,2:22:38.920
So this cat Gladstone was[br]named after this dude, William

2:22:38.920,2:22:43.550
Gladstone, who was an[br]important British politician.

2:22:43.550,2:22:45.760
I'm sure he's not the[br]only thing out there

2:22:45.760,2:22:48.970
in the universe that's named[br]after Gladstone, right?

2:22:48.970,2:22:52.120
I mean there has got[br]to be, I don't know,

2:22:52.120,2:22:54.790
park benches,[br]planets, asteroids,

2:22:54.790,2:22:59.590
something other than the[br]cat, named after this guy.

2:22:59.590,2:23:04.030
So we can ask Wikidata[br]to tell us all the things

2:23:04.030,2:23:06.850
that, you know, without[br]saying instance of something.

2:23:06.850,2:23:10.960
Like, I don't know, anything[br]named after William Gladstone.

2:23:10.960,2:23:12.760
So how do I do that?

2:23:12.760,2:23:15.310
Same principle.

2:23:15.310,2:23:19.850
Instead of asking about the[br]property instance of, property

2:23:19.850,2:23:25.360
31, instead of that, I[br]will ask about the property

2:23:25.360,2:23:26.860
named after--

2:23:26.860,2:23:29.120
sorry, named after--

2:23:29.120,2:23:30.830
I don't need to[br]remember the number.

2:23:30.830,2:23:32.240
I have auto-complete.

2:23:32.240,2:23:35.360
Named after is property 138.

2:23:35.360,2:23:37.430
And I want anything[br]at all that is

2:23:37.430,2:23:42.080
named after this person,[br]William Gladstone.

2:23:42.080,2:23:43.850
Here we go.

2:23:43.850,2:23:45.860
Which is 160852.

2:23:45.860,2:23:46.820
Whatever.

2:23:46.820,2:23:48.230
OK.

2:23:48.230,2:23:50.510
You notice I removed[br]instance of cat.

2:23:50.510,2:23:52.040
I remove the male creature.

2:23:52.040,2:23:55.130
I'm only asking,[br]get me all the items

2:23:55.130,2:23:58.940
that are somehow named after[br]that particular politician.

2:23:58.940,2:24:00.920
And I run the query,[br]and it turns out

2:24:00.920,2:24:05.007
the Wikidata knows[br]about three such things.

2:24:05.007,2:24:06.590
Does that mean that's[br]the only-- these

2:24:06.590,2:24:08.881
are the only three things[br]named after him in the world?

2:24:08.881,2:24:09.939
Of course not.

2:24:09.939,2:24:12.230
But these are the only three[br]items that are in Wikidata

2:24:12.230,2:24:17.720
and explicitly have the[br]property named after Gladstone.

2:24:17.720,2:24:20.150
For all I know, there[br]may be a village

2:24:20.150,2:24:23.600
in England called Gladstone[br]named after this person.

2:24:23.600,2:24:27.410
But if nobody added the[br]property, named after, linking

2:24:27.410,2:24:30.950
to the person, he wouldn't show[br]up in the results to my query.

2:24:30.950,2:24:33.750
So Wikidata knows about[br]three such things.

2:24:33.750,2:24:36.110
One of them is something[br]called the Gladstone Professor

2:24:36.110,2:24:37.360
of Government.

2:24:37.360,2:24:40.370
I can click through and see[br]that it's a chair at Oxford

2:24:40.370,2:24:41.180
University, right?

2:24:41.180,2:24:43.470
So it's a position.

2:24:43.470,2:24:49.520
And another is the William[br]Gladstone school number 18.

2:24:49.520,2:24:51.470
William Gladstone[br]school number 18.

2:24:51.470,2:24:52.900
Where is that?

2:24:52.900,2:24:55.380
That is in Sofia, Bulgaria.

2:24:55.380,2:24:56.470
Again.

2:24:56.470,2:24:59.000
All right, so that's a[br]particular school in Bulgaria

2:24:59.000,2:25:02.720
named after William Gladstone.

2:25:02.720,2:25:07.220
And finally, the third[br]result is, of course, our pal

2:25:07.220,2:25:09.800
Gladstone the Cheif Mouser.

2:25:09.800,2:25:12.674
If I click through,[br]that's the cat.

2:25:12.674,2:25:14.090
All right, so that[br]was an example.

2:25:14.090,2:25:15.700
I mean, you saw how easy it was.

2:25:15.700,2:25:18.980
I just named the property and[br]the value that I care about,

2:25:18.980,2:25:21.420
and I get the results.

2:25:21.420,2:25:23.289
Again, I mean, it's[br]kind of a silly example,

2:25:23.289,2:25:24.080
but think about it.

2:25:24.080,2:25:27.570
This is-- how else can[br]you answer that question?

2:25:27.570,2:25:30.470
There's no reference desk,[br]even at a great University

2:25:30.470,2:25:34.250
of Oxford, where you can[br]walk in and say, give me

2:25:34.250,2:25:37.470
a list of things[br]named after Gladstone.

2:25:37.470,2:25:40.590
There's no easy way to[br]answer that unless you happen

2:25:40.590,2:25:44.520
to have a very large[br]structured and linked

2:25:44.520,2:25:48.130
data store, like Wikidata.

2:25:48.130,2:25:50.560
All right, so that[br]was a silly example.

2:25:50.560,2:25:51.280
Let's take some--

2:25:51.280,2:25:53.113
AUDIENCE: There's a[br]bunch of stuff on there.

2:25:53.113,2:25:54.446
ASAF: Oh, OK.

2:25:54.446,2:25:57.430
AUDIENCE: Can you show[br]easy query on the video?

2:25:57.430,2:26:02.260
And somebody needs to know[br]how to just do property

2:26:02.260,2:26:05.750
exists without giving[br]a specific value.

2:26:05.750,2:26:11.030
And then once you show easy[br]query you reload the page and--

2:26:11.030,2:26:13.240
ASAF: I don't know easy query.

2:26:13.240,2:26:15.670
So is that a gadget?

2:26:15.670,2:26:17.110
I don't know what easy query is.

2:26:17.110,2:26:19.870
I don't use it.

2:26:19.870,2:26:24.760
So someone can maybe[br]send a link or something?

2:26:24.760,2:26:26.100
Oh it is a gadget.

2:26:26.100,2:26:27.100
I don't have it enabled.

2:26:31.610,2:26:32.480
That is nice.

2:26:32.480,2:26:42.080
So now, what I just did by hand,[br]by formulating the query named

2:26:42.080,2:26:45.200
after Gladstone--

2:26:45.200,2:26:48.390
I guess this is the--

2:26:48.390,2:26:48.960
Is it?

2:26:53.000,2:26:53.720
Yeah.

2:26:53.720,2:26:56.050
So this-- I just[br]clicked the three--

2:26:56.050,2:26:57.470
the ellipsis here.

2:26:57.470,2:26:58.460
Right after the name.

2:26:58.460,2:26:59.630
You see this?

2:26:59.630,2:27:03.050
This was just added by[br]enabling easy query,

2:27:03.050,2:27:04.640
which I just learned about.

2:27:04.640,2:27:07.640
So you just click this[br]and it auto-magically

2:27:07.640,2:27:09.620
made this kind of trivial query.

2:27:09.620,2:27:12.380
Of course, if I want a more[br]complicated query like,

2:27:12.380,2:27:14.510
I don't know, give me[br]all the things that

2:27:14.510,2:27:18.110
are named after Lincoln[br]but are a school,

2:27:18.110,2:27:21.650
I will still need to kind[br]of edit a custom query.

2:27:21.650,2:27:23.450
But this is a super[br]easy and very nice

2:27:23.450,2:27:28.620
way of just doing a very super[br]quick query for exactly this.

2:27:28.620,2:27:29.120
Right?

2:27:29.120,2:27:33.410
Like. what other items have[br]exactly this property and value

2:27:33.410,2:27:35.720
named after William Gladstone?

2:27:35.720,2:27:38.750
So, thank you to whoever[br]made this suggestion

2:27:38.750,2:27:42.140
to demonstrate that, and[br]I'm glad I learned something

2:27:42.140,2:27:45.230
too today.

2:27:45.230,2:27:48.590
Let's move to[br]another sample query.

2:27:48.590,2:27:50.360
Here's a fun example.

2:27:50.360,2:27:56.910
Popular surnames among[br]fictional characters.

2:27:56.910,2:27:58.650
Think about that for a second.

2:27:58.650,2:28:03.030
Popular surnames among[br]fictional characters.

2:28:03.030,2:28:06.510
So we're asking Wikidata[br]to go through all

2:28:06.510,2:28:10.120
the fictional[br]characters you know,

2:28:10.120,2:28:13.510
and of those look through[br]their surnames, group

2:28:13.510,2:28:15.910
them so that you can count[br]them, the repetitions

2:28:15.910,2:28:18.460
of the surnames,[br]and give me the most

2:28:18.460,2:28:21.550
popular surnames among them.

2:28:21.550,2:28:26.280
Additionally, I want you to[br]awesomely present the results

2:28:26.280,2:28:28.020
as a bubble chart.

2:28:28.020,2:28:29.220
Oh, yeah.

2:28:29.220,2:28:31.050
Wikidata can do that.

2:28:31.050,2:28:34.420
And I run the query.

2:28:34.420,2:28:36.750
And check it out.

2:28:36.750,2:28:41.130
The most popular names[br]among fictional characters

2:28:41.130,2:28:45.780
we can say that knows about are[br]Joan, Smith, Taylor, et cetera.

2:28:45.780,2:28:48.450
I mean for all we know,[br]the most popular name

2:28:48.450,2:28:50.770
among fictional characters[br]actually in the world

2:28:50.770,2:28:52.350
may be Wu.

2:28:52.350,2:28:54.790
Or something in Chinese[br]for all we know.

2:28:54.790,2:28:57.930
But if that has not been[br]modeled in Wikidata,

2:28:57.930,2:29:01.020
we're not going to get that.

2:29:01.020,2:29:03.540
So Taylor, Smith,[br]Jones, Williams,

2:29:03.540,2:29:06.870
seem to be the[br]most popular names.

2:29:06.870,2:29:08.400
And again, I could limit this.

2:29:08.400,2:29:11.520
I could make the[br]same query but add,

2:29:11.520,2:29:14.250
only among works whose[br]original language

2:29:14.250,2:29:19.020
was Italian, for example, to get[br]more interesting results if I

2:29:19.020,2:29:21.480
only care about[br]Italian literature.

2:29:21.480,2:29:24.720
But this is an example of[br]how I got awesome bubble

2:29:24.720,2:29:28.170
charts for free, and[br]I can just plug this

2:29:28.170,2:29:30.900
into an awesome[br]presentation that I make.

2:29:30.900,2:29:34.500
Of course I can still[br]look at the raw table.

2:29:34.500,2:29:37.940
So the query still resulted[br]in a bunch of data, right?

2:29:37.940,2:29:42.480
So Smith repeats 41 times,[br]Jones 38 times, Taylor 34 times,

2:29:42.480,2:29:43.750
et cetera, et cetera.

2:29:43.750,2:29:48.960
And down that list.

2:29:48.960,2:29:52.320
And I could, again, I could[br]export this into a file

2:29:52.320,2:29:56.100
and load it up in a spreadsheet,[br]and do additional processing

2:29:56.100,2:29:56.670
on it.

2:29:56.670,2:29:58.560
I can link to it.

2:29:58.560,2:30:02.530
I can do all kinds of[br]awesome things with it.

2:30:02.530,2:30:05.250
So that's another awesome query.

2:30:05.250,2:30:08.460
We don't have to go into[br]every line by line analysis

2:30:08.460,2:30:11.670
here of why this[br]works the way it does.

2:30:11.670,2:30:15.840
I want to show you some[br]other queries first.

2:30:15.840,2:30:22.470
Let's look at-- this is just[br]fun, overall causes of death.

2:30:22.470,2:30:24.870
Again a bubble[br]chart just looking

2:30:24.870,2:30:28.260
at people who died[br]of things, and have

2:30:28.260,2:30:30.760
a cause of death listed.

2:30:30.760,2:30:34.380
And we learn that the most[br]commonly listed cause of death

2:30:34.380,2:30:40.350
is myocardial infarction,[br]pneumonitis, cerebral vascular,

2:30:40.350,2:30:42.620
lung cancer, et[br]cetera, et cetera.

2:30:42.620,2:30:44.850
And again, in a bubble chart.

2:30:44.850,2:30:49.670
And so how does that work?

2:30:49.670,2:30:53.050
So just very briefly, the[br]important parts of this query

2:30:53.050,2:30:59.150
are I'm looking for something,[br]for some person, who

2:30:59.150,2:31:04.240
is instance of 31, instance[br]of Q5, which is human.

2:31:04.240,2:31:05.390
So a human.

2:31:05.390,2:31:07.130
Again, just to kind[br]of limit the query.

2:31:07.130,2:31:11.330
I'm not interested in[br]books or mountains.

2:31:11.330,2:31:14.420
I'm looking for humans[br]who have that same person,

2:31:14.420,2:31:21.150
that same variable PID,[br]should have a 509, meaning--

2:31:21.150,2:31:22.412
Hello.

2:31:22.412,2:31:24.620
Why don't I have the--

2:31:24.620,2:31:25.120
Yeah.

2:31:25.120,2:31:28.480
A 509, which is cause of death.

2:31:28.480,2:31:31.540
And that cause of death[br]is another variable,

2:31:31.540,2:31:32.930
that I'm calling CID.

2:31:32.930,2:31:35.410
Now, previously[br]we were saying you

2:31:35.410,2:31:36.850
know I want things[br]that are named

2:31:36.850,2:31:39.550
after Gladstone specifically.

2:31:39.550,2:31:42.000
Only things that have[br]that particular value.

2:31:42.000,2:31:44.320
Here I'm saying I'm[br]looking for things

2:31:44.320,2:31:47.110
that have some cause of death.

2:31:47.110,2:31:48.760
Not a specific one.

2:31:48.760,2:31:50.260
I just wanted to[br]get everything that

2:31:50.260,2:31:54.880
has a statement with some[br]value about property 509

2:31:54.880,2:31:56.530
cause of death.

2:31:56.530,2:31:57.940
OK?

2:31:57.940,2:32:04.410
And then this other bit of[br]magic here, the group by,

2:32:04.410,2:32:07.870
tells Wikidata I'm not[br]actually interested

2:32:07.870,2:32:09.100
in every individual thing.

2:32:09.100,2:32:12.310
I want you to group those[br]causes, and then count them

2:32:12.310,2:32:14.230
and give me the top ones.

2:32:14.230,2:32:15.523
So that's how this query works.

2:32:20.550,2:32:22.320
Here's that query I promised.

2:32:22.320,2:32:26.460
Painters whose fathers[br]were also painters.

2:32:26.460,2:32:28.630
I can only think of a couple.

2:32:28.630,2:32:31.890
I mean, Monet and Vogel.

2:32:31.890,2:32:34.800
But I'm sure Wikidata[br]knows many more.

2:32:34.800,2:32:38.620
So let's run this query.

2:32:38.620,2:32:40.270
And I have 100 results.

2:32:40.270,2:32:43.120
By the way, I have limited[br]it to 100 results just

2:32:43.120,2:32:44.650
to keep it kind of snappy.

2:32:44.650,2:32:47.530
But actually, we could[br]maybe try removing the limit

2:32:47.530,2:32:50.170
and see if Wikidata[br]could tell us

2:32:50.170,2:32:53.890
the total number in Wikidata.

2:32:53.890,2:32:55.120
Yeah, that wasn't too bad.

2:32:55.120,2:32:58.400
So 1,270 results.

2:32:58.400,2:32:59.140
OK.

2:32:59.140,2:33:04.150
Wikidata, already at this[br]early date and it's progress,

2:33:04.150,2:33:07.540
already knows about[br]more than 1,200 painters

2:33:07.540,2:33:10.980
who are sons of painters.

2:33:10.980,2:33:16.140
Sons of male painters, like[br]their father is a painter.

2:33:16.140,2:33:18.120
There may be[br]additional painters who

2:33:18.120,2:33:21.390
are sons of female painters[br]not included in this query.

2:33:21.390,2:33:24.990
Again, always remember what[br]exactly you are asking.

2:33:24.990,2:33:27.840
In this query I was[br]asking about the father.

2:33:27.840,2:33:30.330
I'm leaving out any[br]possible painters who

2:33:30.330,2:33:32.720
are sons of mother painters.

2:33:32.720,2:33:33.390
OK?

2:33:33.390,2:33:35.250
So how does this work?

2:33:35.250,2:33:39.630
I'm asking for the painter[br]along with the human label,

2:33:39.630,2:33:42.630
and the father along[br]with the human label.

2:33:42.630,2:33:47.610
So Michel Monet is the[br]son of Claude Monet.

2:33:47.610,2:33:54.180
And Domenico Tintoretto is the[br]son of the famous Tintoretto

2:33:54.180,2:33:57.210
whose label, you know, is just[br]Tintoretto like Michelangelo.

2:33:57.210,2:33:59.960
You know, you don't always[br]have to have the full name

2:33:59.960,2:34:02.420
in the common label.

2:34:02.420,2:34:07.010
Paloma Picasso is the[br]daughter of Pablo Picasso.

2:34:07.010,2:34:07.510
OK.

2:34:07.510,2:34:11.040
So Wikidata knows about[br]all these results.

2:34:11.040,2:34:14.610
Of course Holbein the Younger[br]son of Holbein the Elder.

2:34:14.610,2:34:15.760
And how did we get there?

2:34:15.760,2:34:20.860
Well we asked Wikidata[br]to look for something,

2:34:20.860,2:34:26.820
let's call it painter, which[br]has 106, which is occupation,

2:34:26.820,2:34:31.100
with a value painter.

2:34:31.100,2:34:31.600
Right?

2:34:31.600,2:34:35.310
This unwieldy number[br]1028181, that's painter.

2:34:35.310,2:34:40.250
So I'm asking for any item[br]that has occupation painter.

2:34:40.250,2:34:43.300
And let's call[br]that item painter.

2:34:43.300,2:34:49.770
I also want that painter to have[br]a property 22, which is father.

2:34:49.770,2:34:50.850
OK.

2:34:50.850,2:34:52.350
Father.

2:34:52.350,2:34:55.140
And I want it to[br]have some value.

2:34:55.140,2:34:58.770
OK, I'm putting it into[br]another variable called father.

2:34:58.770,2:35:01.320
I could have called[br]it, you know, frog.

2:35:01.320,2:35:04.230
That doesn't change[br]anything, just to be clear.

2:35:04.230,2:35:06.630
What matters is that this[br]is the property father.

2:35:06.630,2:35:10.320
I could have called[br]it anything I want.

2:35:10.320,2:35:13.590
So, and then, I have[br]a third condition.

2:35:13.590,2:35:18.010
That the father, like whatever[br]it says here in property 22,

2:35:18.010,2:35:22.590
I want that father to have[br]himself a property 106

2:35:22.590,2:35:27.750
occupation with a value painter.

2:35:27.750,2:35:28.730
OK?

2:35:28.730,2:35:30.800
These conditions[br]combined to give me

2:35:30.800,2:35:36.080
a list of people who have[br]a father and that father

2:35:36.080,2:35:37.850
has occupation painter as well.

2:35:37.850,2:35:40.550
Of course, if I suddenly,[br]or if you suddenly,

2:35:40.550,2:35:44.480
are consumed by[br]curiosity to know

2:35:44.480,2:35:51.344
who are some politicians[br]who are sons of carpenters?

2:35:51.344,2:35:52.760
You could just[br]change that, right?

2:35:52.760,2:35:56.700
Change the first value[br]from painter to politician.

2:35:56.700,2:36:02.624
Change the third line's value[br]from painter to carpenter.

2:36:02.624,2:36:04.040
Maybe that list[br]will be very short

2:36:04.040,2:36:06.680
because carpenters don't[br]tend to be notable,

2:36:06.680,2:36:08.910
so they wouldn't be[br]represented on Wikidata.

2:36:08.910,2:36:11.990
That's why this works relatively[br]well with painters, right?

2:36:11.990,2:36:14.420
Because most of[br]them are notable.

2:36:14.420,2:36:16.370
But generally you[br]could do that, right?

2:36:16.370,2:36:18.500
That's an example of[br]how you can take a query

2:36:18.500,2:36:22.340
and just replace one of those[br]values, or even the language.

2:36:22.340,2:36:26.840
So again, I could ask[br]for these same painters.

2:36:26.840,2:36:27.650
It's limited again.

2:36:27.650,2:36:31.190
These same painters,[br]but with Arabic labels.

2:36:31.190,2:36:34.880
Same query, but I have Arabic[br]labels for these painters.

2:36:34.880,2:36:37.250
And of course where[br]there is no Arabic label

2:36:37.250,2:36:40.360
I get the queue number.

2:36:40.360,2:36:40.860
OK?

2:36:40.860,2:36:43.650
So that's that query[br]that I promised you,

2:36:43.650,2:36:47.670
painters who sons of painters[br]can be done by Wikidata

2:36:47.670,2:36:49.830
in under one second.

2:36:49.830,2:36:51.480
How awesome is that?

2:36:51.480,2:36:52.950
We can also get some statistics.

2:36:52.950,2:36:55.920
So how about counting[br]total articles

2:36:55.920,2:36:59.740
in a given wiki by gender.

2:36:59.740,2:37:02.070
This is what we call[br]the content gender

2:37:02.070,2:37:06.900
gap, as distinct from the[br]participation gender gap.

2:37:06.900,2:37:10.276
This is the gender gap in[br]what we cover on Wikipedia.

2:37:10.276,2:37:11.400
So let's take one of these.

2:37:16.380,2:37:17.630
So this is a query.

2:37:17.630,2:37:23.130
Articles about women in[br]some given Wikipedia.

2:37:23.130,2:37:23.660
All right.

2:37:23.660,2:37:25.799
So let's take--

2:37:25.799,2:37:26.340
I don't know.

2:37:26.340,2:37:30.240
Let's take the Tamil Wikipedia.

2:37:30.240,2:37:32.460
That's language code TA.

2:37:32.460,2:37:34.950
So I just put TA here.

2:37:34.950,2:37:38.850
And I click Run, and[br]I get this count.

2:37:38.850,2:37:39.960
That's all I wanted.

2:37:39.960,2:37:41.720
I'm not actually[br]interested in the items,

2:37:41.720,2:37:44.962
like in the list of women[br]on the Tamil Wikipedia.

2:37:44.962,2:37:45.920
I just want the number.

2:37:45.920,2:37:48.510
So I selected the count here.

2:37:48.510,2:37:52.610
And this number[br]turns out to be 2159.

2:37:52.610,2:37:57.300
So there are 2000[br]articles about women

2:37:57.300,2:38:02.350
the Tamil Wikipedia that[br]Wikidata knows to be female.

2:38:02.350,2:38:02.850
Right?

2:38:02.850,2:38:05.730
I'm asking about the gender[br]field, property 21 again.

2:38:05.730,2:38:08.900
Remember, if there's some[br]article about a woman in Tamil

2:38:08.900,2:38:12.090
Wikipedia, but wiki[br]data doesn't have

2:38:12.090,2:38:14.460
a statement about the[br]gender, that person

2:38:14.460,2:38:15.640
will not be counted here.

2:38:15.640,2:38:18.240
So again, be careful[br]about kind of stating

2:38:18.240,2:38:22.800
that is exactly the number[br]of women articles on Tamil

2:38:22.800,2:38:23.340
Wikipedia.

2:38:23.340,2:38:24.600
That's probably not true.

2:38:24.600,2:38:27.560
I'm sure some of those[br]articles are missing

2:38:27.560,2:38:30.740
a sex or gender or property.

2:38:30.740,2:38:33.150
But for raw statistics,[br]that's probably good,

2:38:33.150,2:38:35.700
because some men are also[br]missing the sex or gender

2:38:35.700,2:38:37.620
statistic property.

2:38:37.620,2:38:41.820
So we could take the[br]same query for men.

2:38:41.820,2:38:43.170
It's essentially the exact same.

2:38:43.170,2:38:48.840
It just has this unwieldy[br]number for males, 6581097.

2:38:48.840,2:38:52.710
I can change this language[br]code again to TA for Tamil.

2:38:52.710,2:38:58.880
And how many men are covered[br]on Tamil Wikipedia 14,649.

2:38:58.880,2:38:59.610
OK.

2:38:59.610,2:39:06.880
So women, 2,100, men,[br]about seven times as many.

2:39:06.880,2:39:07.380
Right?

2:39:07.380,2:39:12.300
So that's the approximate[br]size of the content gender

2:39:12.300,2:39:14.610
gap on Tamil Wikipedia.

2:39:14.610,2:39:18.850
And again, I can complicate[br]this query as much as I want.

2:39:18.850,2:39:21.390
For example, I can[br]try and find out

2:39:21.390,2:39:30.390
if this gender gap is wider[br]or narrower among musicians,

2:39:30.390,2:39:31.350
just as an example.

2:39:31.350,2:39:35.850
I could just add a line here[br]that says occupation musician,

2:39:35.850,2:39:37.890
and then I'm only[br]counting articles

2:39:37.890,2:39:41.190
on Tamil Wikipedia about[br]musicians who are female

2:39:41.190,2:39:43.190
versus articles[br]on Tamil Wikipedia

2:39:43.190,2:39:45.030
about musicians who are male.

2:39:45.030,2:39:47.890
And I can kind of[br]compare the gender--

2:39:47.890,2:39:53.820
the content gender gap across[br]occupations on Tamil Wikipedia.

2:39:53.820,2:39:56.030
Do you see the[br]important point here?

2:39:56.030,2:39:58.490
Is that this is not just[br]kind of a one purpose query.

2:39:58.490,2:40:01.250
I can just with a single[br]additional conditional suddenly

2:40:01.250,2:40:04.370
make it a much more interesting[br]query, because I break it down

2:40:04.370,2:40:05.540
by occupation.

2:40:05.540,2:40:07.810
Or I break it down by century.

2:40:07.810,2:40:12.530
Do we have more of the coverage[br]gap in 19th century people

2:40:12.530,2:40:13.940
than in 21st century people?

2:40:13.940,2:40:15.560
I mean, I sure hope so, right?

2:40:15.560,2:40:18.480
The patriarchy is[br]weakening somewhat.

2:40:18.480,2:40:21.830
So I wouldn't be surprised if[br]there are many more notable men

2:40:21.830,2:40:23.430
covered about the 19th century.

2:40:23.430,2:40:25.784
But if we are also covering--

2:40:25.784,2:40:27.200
I mean it's the[br]gender gap is just

2:40:27.200,2:40:29.540
as wide for 21st century[br]people, that would

2:40:29.540,2:40:30.800
be a little disappointing.

2:40:30.800,2:40:35.870
Again that's something I[br]can fairly easily find out

2:40:35.870,2:40:38.980
on Wikidata query.

2:40:38.980,2:40:41.500
Any questions so far, or[br]are you just sharing links?

2:40:41.500,2:40:43.160
AUDIENCE: Yep there is one.

2:40:43.160,2:40:47.480
So somebody is wondering if you[br]can demonstrate, or at least

2:40:47.480,2:40:50.420
give a short answer of the[br]latter of this question.

2:40:50.420,2:40:52.530
Is it possible using[br]in Wikidata SPARQL

2:40:52.530,2:40:55.520
to find specific[br]Wikidata articles, e.g.

2:40:55.520,2:40:59.060
featured articles, of a[br]certain language which do not

2:40:59.060,2:41:01.160
exist in another language.

2:41:01.160,2:41:03.770
I know it is possible[br]to find category based

2:41:03.770,2:41:05.820
results using a PET scan tool.

2:41:05.820,2:41:09.110
But can we specify[br]that by selecting e.g.

2:41:09.110,2:41:10.055
featured articles?

2:41:10.055,2:41:11.390
ASAF BARTOV: Yes.

2:41:11.390,2:41:12.600
Excellent question.

2:41:12.600,2:41:14.120
It is possible, indeed.

2:41:14.120,2:41:17.570
And I will demonstrate[br]one such query.

2:41:17.570,2:41:19.190
Another query that[br]I already mentioned

2:41:19.190,2:41:24.840
largest cities in the[br]world with a female mayor.

2:41:24.840,2:41:29.190
This query-- let's[br]close some of these tabs

2:41:29.190,2:41:30.315
before my browser chokes.

2:41:33.600,2:41:36.840
So this query lists[br]the major world cities

2:41:36.840,2:41:39.120
run by women currently.

2:41:39.120,2:41:45.650
And the answer is Mumbai, Mexico[br]City, Tokyo, bunch of others.

2:41:49.470,2:41:52.371
And wait-- that's not it at all.

2:41:52.371,2:41:53.370
I clicked the wrong one.

2:41:53.370,2:41:55.050
That's the map of paintings.

2:41:55.050,2:41:55.800
OK.

2:41:55.800,2:41:57.370
Let's demonstrate[br]that for a second.

2:41:57.370,2:41:59.520
So this is the map[br]of all paintings

2:41:59.520,2:42:03.870
for which we know a location[br]with the count per location.

2:42:03.870,2:42:07.770
And the results are[br]awesomely presented on a map.

2:42:07.770,2:42:08.830
OK.

2:42:08.830,2:42:12.420
Again, under the hood this is[br]a table, of course, of results.

2:42:12.420,2:42:15.660
But, awesomely, I can[br]browse it as a map.

2:42:15.660,2:42:20.320
So here is a map of the[br]world with all the paintings

2:42:20.320,2:42:22.060
that Wikidata knows about.

2:42:22.060,2:42:23.920
Not just knows[br]about the paintings,

2:42:23.920,2:42:28.180
but knows about their[br]location in a museum.

2:42:28.180,2:42:30.670
Not surprisingly[br]Europe is much better

2:42:30.670,2:42:35.540
covered than Russia or Africa.

2:42:35.540,2:42:40.150
There is a huge gap in[br]contribution to Wikidata

2:42:40.150,2:42:41.740
from these countries.

2:42:41.740,2:42:43.780
And some of it can be fixed.

2:42:43.780,2:42:47.740
And of course there is much more[br]documentation, and much more

2:42:47.740,2:42:50.260
art in Europe.

2:42:50.260,2:42:54.280
But if we zoom in, I[br]don't know, Rome probably

2:42:54.280,2:42:55.900
has a few paintings.

2:42:55.900,2:42:56.400
Right?

2:43:00.080,2:43:02.288
Hello.

2:43:02.288,2:43:04.200
Sorry.

2:43:04.200,2:43:09.780
It's-- Yes.

2:43:09.780,2:43:13.290
Vatican City sounds[br]like a good bet, right?

2:43:13.290,2:43:14.290
I can zoom in here.

2:43:14.290,2:43:16.290
And I can just click[br]one of these dots

2:43:16.290,2:43:21.400
and see in this point[br]there are two paintings.

2:43:21.400,2:43:25.270
And in this one there is one[br]and it's the Archbasilica

2:43:25.270,2:43:27.460
of St. John Lateran.

2:43:27.460,2:43:31.060
Let's see, this is the[br]actual St. Peter, right?

2:43:31.060,2:43:33.650
Sistine Chapel has 23 paintings.

2:43:33.650,2:43:34.330
What?

2:43:34.330,2:43:36.670
The Sistine Chapel has way[br]more than 23 paintings.

2:43:36.670,2:43:40.330
Correct, but 23 of them[br]are documented on Wikidata.

2:43:40.330,2:43:43.330
Have their own item[br]for the painting, not

2:43:43.330,2:43:45.280
the Sistine Chapel,[br]the painting has

2:43:45.280,2:43:49.540
an item that lists its[br]being in the Sistine Chapel.

2:43:49.540,2:43:50.950
There are 23 of those.

2:43:50.950,2:43:52.270
OK.

2:43:52.270,2:43:54.310
There is definitely[br]room to document

2:43:54.310,2:43:57.040
the rest of the artworks[br]in the Sistine Chapel.

2:43:57.040,2:43:59.740
So, again, this is just[br]not the kind of query

2:43:59.740,2:44:03.330
you were able to[br]make before Wikidata,

2:44:03.330,2:44:07.750
and it's a fairly simple[br]query, as you can see.

2:44:07.750,2:44:13.020
There are examples using[br]maps like airports within 100

2:44:13.020,2:44:15.040
kilometers of Berlin.

2:44:15.040,2:44:18.310
Again using the coordinates[br]as a useful data point.

2:44:18.310,2:44:21.880
And here is a map showing me[br]only airports within a 100

2:44:21.880,2:44:25.990
kilometer radius from Berlin.

2:44:25.990,2:44:29.140
But I wanted to show[br]you the mayors query.

2:44:29.140,2:44:34.510
Let's click the-- oh I just[br]have the wrong link here.

2:44:34.510,2:44:41.040
But I can still find it[br]here by typing mayor.

2:44:41.040,2:44:44.590
Here we go, largest[br]cities with female mayor.

2:44:44.590,2:44:47.230
So this is a slightly[br]more complicated query.

2:44:47.230,2:44:53.010
But if I run it, I get the top[br]10, because I set limit to 10.

2:44:53.010,2:44:54.820
I get the top 10[br]cities in the world,

2:44:54.820,2:44:59.710
by population, size that[br]are currently run by women.

2:44:59.710,2:45:03.490
Tokyo, Mumbai, Yokohama,[br]Caracas, et cetera.

2:45:03.490,2:45:08.080
And one interesting thing that[br]you may want to notice here

2:45:08.080,2:45:10.690
is that I'm asking for cities.

2:45:10.690,2:45:13.660
I mean items, that[br]are instance of city.

2:45:13.660,2:45:16.420
And that have a[br]head of government,

2:45:16.420,2:45:18.640
that have some[br]statement about who

2:45:18.640,2:45:28.440
is in charge, and that statement[br]has sex that's listed up here

2:45:28.440,2:45:29.886
as female.

2:45:29.886,2:45:31.510
Don't worry about[br]the syntax right now.

2:45:31.510,2:45:34.590
I just want to show you[br]some specific angle here.

2:45:34.590,2:45:37.920
And I'm further[br]filtering these results.

2:45:37.920,2:45:45.400
I only want those items where[br]there is not the property

2:45:45.400,2:45:48.630
and the qualifier, end time.

2:45:48.630,2:45:50.390
Why is that important?

2:45:50.390,2:45:56.530
Because if a city once[br]had a female mayor,

2:45:56.530,2:45:59.890
but that mayor is not the mayor[br]anymore, because mayors change,

2:45:59.890,2:46:01.600
I don't want them in this query.

2:46:01.600,2:46:04.990
I want to query of[br]cities currently having

2:46:04.990,2:46:05.680
a female mayor.

2:46:05.680,2:46:07.990
And of course Wikidata[br]may have historical data

2:46:07.990,2:46:09.880
with start and[br]end time, as we've

2:46:09.880,2:46:14.530
seen, that documents this[br]person was the mayor of Tokyo

2:46:14.530,2:46:17.170
or San Francisco[br]between these years.

2:46:17.170,2:46:18.820
But if there is no[br]end times that means

2:46:18.820,2:46:21.520
they are currently the mayor.

2:46:21.520,2:46:24.490
So that's an example of[br]asking about a qualifier

2:46:24.490,2:46:28.180
of a statement, to again, to get[br]the results we actually want.

2:46:28.180,2:46:31.630
If we want current mayors it's[br]important to put this filter.

2:46:31.630,2:46:35.365
If we don't, we will get[br]historical female mayors

2:46:35.365,2:46:35.865
as well.

2:46:39.920,2:46:40.490
All right.

2:46:40.490,2:46:45.380
So these are some[br]example queries.

2:46:45.380,2:46:49.085
Questions about that?

2:46:51.620,2:46:53.030
Oh, the featured[br]article example.

2:46:58.280,2:47:01.700
So let's look at that.

2:47:07.050,2:47:12.660
So I have prepared[br]such a query recently.

2:47:12.660,2:47:15.300
Here we go.

2:47:15.300,2:47:18.570
So this is a query.

2:47:18.570,2:47:20.472
I just saved it here[br]on my user page.

2:47:20.472,2:47:21.930
I mean, this is[br]not Wikidata query.

2:47:21.930,2:47:25.390
This is just a meta page[br]containing the query usefully.

2:47:28.260,2:47:33.800
And let's run this.

2:47:33.800,2:47:38.030
So this query, it's actually[br]not very complicated.

2:47:38.030,2:47:40.030
It's just has a long[br]list of countries,

2:47:40.030,2:47:42.170
because I'm asking[br]about African countries.

2:47:42.170,2:47:42.670
OK.

2:47:42.670,2:47:45.010
I'm looking for human[br]females from one

2:47:45.010,2:47:51.060
of these countries that[br]have an article in English.

2:47:51.060,2:47:53.010
That's what this line means.

2:47:53.010,2:47:55.620
But not in French.

2:47:55.620,2:47:57.570
That's what this part means.

2:47:57.570,2:47:59.170
OK.

2:47:59.170,2:48:01.720
This part, these[br]two lines together.

2:48:01.720,2:48:03.190
But not in French.

2:48:03.190,2:48:05.920
And this is what's[br]called a badge.

2:48:05.920,2:48:09.430
That's Wikidata's concept of[br]good and featured articles.

2:48:09.430,2:48:10.600
It's called a badge.

2:48:10.600,2:48:16.500
So I want them to have some[br]badge on English Wikipedia.

2:48:16.500,2:48:17.000
OK?

2:48:17.000,2:48:22.250
So again, this query is[br]asking for the top 100 women

2:48:22.250,2:48:26.150
from Africa who are documented[br]on English Wikipedia,

2:48:26.150,2:48:28.730
in a featured or[br]good article status.

2:48:28.730,2:48:30.660
But not on French Wikipedia.

2:48:30.660,2:48:33.270
So this is a query that's[br]a to-do query, right?

2:48:33.270,2:48:35.630
That's a query[br]for French editors

2:48:35.630,2:48:40.100
to consider what they might[br]usefully translate or create

2:48:40.100,2:48:41.180
in French.

2:48:41.180,2:48:48.860
And if we run this see[br]we have three results.

2:48:48.860,2:48:50.720
I mean, we have many[br]women from Africa

2:48:50.720,2:48:52.460
covered on English Wikipedia.

2:48:52.460,2:48:57.500
But only three articles[br]have featured or good status

2:48:57.500,2:49:03.460
among those that do not have[br]French Wikipedia coverage.

2:49:03.460,2:49:04.900
Let me rephrase that.

2:49:04.900,2:49:07.990
Among the English Wikipedia[br]articles about African women

2:49:07.990,2:49:11.170
that don't have a[br]French counterpart,

2:49:11.170,2:49:14.520
only three are featured or good.

2:49:14.520,2:49:16.960
OK?

2:49:16.960,2:49:17.640
Do you see this?

2:49:17.640,2:49:19.720
The badge is good article.

2:49:19.720,2:49:23.550
This little incantation[br]here is what allows

2:49:23.550,2:49:25.950
you to ask about the badge.

2:49:25.950,2:49:28.730
This here.

2:49:28.730,2:49:33.420
And, by the way, the slides[br]will be uploaded to commons.

2:49:33.420,2:49:38.708
And we will-- how shall we make[br]it available on the YouTube

2:49:38.708,2:49:39.710
thing as well?

2:49:42.730,2:49:43.230
No, no.

2:49:43.230,2:49:45.870
But, I mean, for people who[br]will later watch this video.

2:49:52.119,2:49:54.160
Oh yeah, we can add it to[br]the YouTube description

2:49:54.160,2:49:55.368
and the comments description.

2:49:55.368,2:49:58.090
So in the-- if you're[br]watching this video later,

2:49:58.090,2:50:00.820
in the description, we will[br]add a link to this query

2:50:00.820,2:50:01.480
specifically.

2:50:01.480,2:50:03.340
Because it's not in[br]the slides right now.

2:50:03.340,2:50:03.910
It will be.

2:50:06.622,2:50:07.980
OK.

2:50:07.980,2:50:10.260
So.

2:50:10.260,2:50:13.590
Questions so far?

2:50:13.590,2:50:14.700
We're almost done.

2:50:14.700,2:50:16.260
We have a few minutes left.

2:50:16.260,2:50:18.090
So questions about queries?

2:50:18.090,2:50:20.130
I mean, I'm sure[br]there's tons of things

2:50:20.130,2:50:21.510
you don't know how to do yet.

2:50:21.510,2:50:24.720
And you maybe you didn't really[br]get the sense for SPARQL.

2:50:24.720,2:50:27.120
It's something you need[br]to really do on your own

2:50:27.120,2:50:28.290
on your computer.

2:50:28.290,2:50:29.465
See how it works.

2:50:29.465,2:50:30.090
Fiddle with it.

2:50:30.090,2:50:30.900
Change something.

2:50:30.900,2:50:33.270
See that it breaks[br]and complains.

2:50:33.270,2:50:37.470
But, very importantly-- oh I[br]had this in the other questions

2:50:37.470,2:50:38.340
slide.

2:50:38.340,2:50:42.480
Remember Wikidata project chat.

2:50:42.480,2:50:45.810
That's kind of the Wikidata[br]equivalent of the village pump.

2:50:45.810,2:50:47.790
It's the page on Wikidata[br]where you can just

2:50:47.790,2:50:49.830
show up and ask a question.

2:50:49.830,2:50:52.290
In my experience, the[br]Wikidata community

2:50:52.290,2:50:55.410
is very nice, very[br]welcoming, and very eager

2:50:55.410,2:51:00.100
to help newer people integrate[br]and learn how to do things.

2:51:00.100,2:51:01.800
There's also an IRC channel.

2:51:01.800,2:51:04.260
If you know what IRC is and[br]how to use it, by all means,

2:51:04.260,2:51:07.890
go to IRC channel Wikidata.

2:51:07.890,2:51:09.330
There's people[br]there all the time,

2:51:09.330,2:51:11.040
and you can just ask a question.

2:51:11.040,2:51:13.245
If you're trying to do a[br]query, and you don't quite

2:51:13.245,2:51:15.870
understand the syntax, or you're[br]not sure how to get the result

2:51:15.870,2:51:16.680
you want.

2:51:16.680,2:51:20.050
There are people there who[br]will gladly help you do that.

2:51:20.050,2:51:22.560
There is also a[br]Wikidata newsletter

2:51:22.560,2:51:25.680
published by the Wikidata team,[br]which is centered in Germany

2:51:25.680,2:51:27.330
and Wikipedia Germany.

2:51:27.330,2:51:31.890
And they send out a newsletter[br]in English with Wikidata news.

2:51:31.890,2:51:33.570
You know, new[br]properties, new items,

2:51:33.570,2:51:34.920
new things in the project.

2:51:34.920,2:51:36.840
But also sample queries.

2:51:36.840,2:51:39.300
So once a week there is[br]kind of an awesome query

2:51:39.300,2:51:43.440
to learn from, if you want[br]to learn that way instead

2:51:43.440,2:51:46.230
of reading like a[br]whole manual on SPARQL.

2:51:46.230,2:51:48.300
So I'm just encouraging[br]you to get help

2:51:48.300,2:51:49.470
in one of those channels.

2:51:49.470,2:51:51.000
Of course you can write to me.

2:51:51.000,2:51:55.920
Just reach out to me and[br]ask me questions as well.

2:51:55.920,2:51:58.860
I hope by now you agree[br]that Wikidata is love,

2:51:58.860,2:52:03.150
and Wikidata data is awesome.

2:52:03.150,2:52:06.480
If there are no questions,[br]we do have a tiny bit of time

2:52:06.480,2:52:11.510
to demonstrate one[br]more tool but that's--

2:52:11.510,2:52:12.010
no?

2:52:12.010,2:52:13.170
No questions.

2:52:13.170,2:52:17.600
OK so let's talk about--

2:52:17.600,2:52:19.100
well, the resonator[br]is kind of nice,

2:52:19.100,2:52:22.890
but it's a little like[br]the article placeholder.

2:52:22.890,2:52:25.530
So this is not Wikidata[br]this is a tool again

2:52:25.530,2:52:26.805
built by Magnus Manske--

2:52:26.805,2:52:29.310
AUDIENCE: There's also one[br]final question to you in case--

2:52:29.310,2:52:29.820
ASAF BARTOV: Oh,[br]there is a question.

2:52:29.820,2:52:30.390
AUDIENCE: Yeah.

2:52:30.390,2:52:32.348
ASAF BARTOV: Which[br]advantages and disadvantages

2:52:32.348,2:52:35.370
to create an item[br]before an article is

2:52:35.370,2:52:37.920
done on English Wikipedia?

2:52:37.920,2:52:42.340
Well, I mean, this example[br]that I just made right.

2:52:42.340,2:52:46.960
I'm reading this book[br]by a notable author.

2:52:46.960,2:52:47.810
OK.

2:52:47.810,2:52:51.400
I want this to[br]exist on Wikidata,

2:52:51.400,2:52:53.320
and to be mentioned[br]on Wikidata, so

2:52:53.320,2:52:56.950
that when people look up[br]that author in Wikidata

2:52:56.950,2:52:59.170
they will know about one[br]of his notable works.

2:52:59.170,2:53:02.470
But I'm not prepared to[br]put in the time investment

2:53:02.470,2:53:05.670
to build a whole article[br]on English Wikipedia.

2:53:05.670,2:53:07.420
Either because I don't[br]have the time, or I

2:53:07.420,2:53:09.040
don't have good sources.

2:53:09.040,2:53:11.560
Or maybe my English[br]is not good enough,

2:53:11.560,2:53:14.980
but it is good enough to just[br]record these very basic facts

2:53:14.980,2:53:17.850
and point to the Library of[br]Congress records et cetera.

2:53:17.850,2:53:20.170
So that it's better[br]than nothing.

2:53:20.170,2:53:23.170
So that's one reason[br]to maybe do it.

2:53:23.170,2:53:26.690
Another reason is to[br]be able to link to it.

2:53:26.690,2:53:30.190
So remember that[br]translator lady already

2:53:30.190,2:53:33.280
had an item on Wikidata, but if[br]she hadn't we could have just

2:53:33.280,2:53:38.560
created a very, very basic[br]rudimentary item about her just

2:53:38.560,2:53:41.740
saying, you know,[br]this name is human.

2:53:41.740,2:53:43.060
Country, Bulgaria.

2:53:43.060,2:53:45.220
Occupation, translator.

2:53:45.220,2:53:48.580
Even just that would have[br]would have been something,

2:53:48.580,2:53:51.610
and would have enabled me[br]to link to this person.

2:53:51.610,2:53:56.860
So these are legitimate reasons[br]to create Wikidata entities

2:53:56.860,2:54:01.510
without, or at least before,[br]creating a Wikipedia article.

2:54:01.510,2:54:02.709
If you are going to create--

2:54:02.709,2:54:04.750
I mean if you're at and[br]edit-a-thon or something,

2:54:04.750,2:54:07.690
and you have come to[br]create Wikipedia articles,

2:54:07.690,2:54:10.660
by all means, first create[br]the Wikipedia article,

2:54:10.660,2:54:13.982
then create the Wikipedia[br]item and link to it.

2:54:17.580,2:54:20.500
I hope that answers[br]the question.

2:54:20.500,2:54:24.940
So the reasonator[br]is simply a kind

2:54:24.940,2:54:31.330
of prettier view of[br]items in Wikidata.

2:54:31.330,2:54:35.980
So you can just type the name[br]of an item or the number.

2:54:35.980,2:54:39.010
Let's pick just a[br]random number, 42.

2:54:39.010,2:54:39.595
Say 42.

2:54:42.770,2:54:45.950
Which happens to[br]be, maybe you've

2:54:45.950,2:54:51.310
heard of this guy,[br]Douglas Adams.

2:54:51.310,2:54:55.490
He happened to have received[br]the queue number 42.

2:54:55.490,2:54:58.760
I'm sure it's a[br]cosmic coincidence

2:54:58.760,2:55:01.460
of infinite improbability.

2:55:01.460,2:55:03.470
And this is a view--

2:55:03.470,2:55:05.690
this is a tool that[br]is not Wikidata.

2:55:05.690,2:55:09.690
It's a tool built on top of[br]Wikidata called resonator.

2:55:09.690,2:55:14.750
And it gives us the information[br]from Q42, that is from the--

2:55:14.750,2:55:18.800
this item in Wikidata, which[br]looks like an item in Wikidata.

2:55:18.800,2:55:21.320
But it gives it to us in a[br]slightly more rational kind

2:55:21.320,2:55:22.430
of lay out.

2:55:22.430,2:55:24.200
It even kind of[br]generates a little bit

2:55:24.200,2:55:27.620
of pseudo article text for us.

2:55:27.620,2:55:30.429
You know, Douglas Adams was[br]a British writer, playwright,

2:55:30.429,2:55:31.970
screenwriter,[br]bla-bla-bla, an author.

2:55:31.970,2:55:35.630
He was born on this date, in[br]this place, to these people.

2:55:35.630,2:55:39.080
He studied at this place[br]between these years.

2:55:39.080,2:55:40.670
That's all machine generated.

2:55:40.670,2:55:42.230
Nobody wrote this text.

2:55:42.230,2:55:46.330
That's all taken from those[br]statements in Wikidata,

2:55:46.330,2:55:51.080
and generates this reasonable[br]reading summary paragraph.

2:55:51.080,2:55:54.140
And then it gives us this[br]little table of relatives.

2:55:54.140,2:55:55.610
It's all taken from Wikidata.

2:55:55.610,2:55:57.740
But as you can see,[br]this is already

2:55:57.740,2:56:02.120
a little more accessible than[br]the essentially arbitrary

2:56:02.120,2:56:05.120
ordering of statements[br]on Wikidata.

2:56:05.120,2:56:06.200
And that's OK.

2:56:06.200,2:56:08.060
I mean, that's[br]kind of by design.

2:56:08.060,2:56:10.100
Wikidata is the platform.

2:56:10.100,2:56:11.960
There is going to[br]be-- there are going

2:56:11.960,2:56:15.680
to be many new applications,[br]and platforms, and tools,

2:56:15.680,2:56:19.010
and visual interfaces[br]on top of Wikidata

2:56:19.010,2:56:23.000
to browse Wikidata in a more[br]friendly or more customized

2:56:23.000,2:56:24.480
ways.

2:56:24.480,2:56:27.080
For example, one of the[br]things that resonator

2:56:27.080,2:56:31.610
does for us is give us pictures[br]and maps and a timeline.

2:56:31.610,2:56:32.960
Check it out this.

2:56:32.960,2:56:38.990
Time line machine generated,[br]just from dates and points

2:56:38.990,2:56:44.090
in time, mentioned in the[br]relatively rich Wikidata

2:56:44.090,2:56:47.200
item about Douglas Adams.

2:56:47.200,2:56:47.700
Right?

2:56:47.700,2:56:50.030
So this timeline, for example[br]again, completely machine

2:56:50.030,2:56:51.140
generated.

2:56:51.140,2:56:53.270
But he was educated[br]between these years,

2:56:53.270,2:56:54.920
so I can put it on the timeline.

2:56:54.920,2:56:57.260
And this is the year he was[br]nominated for a Hugo awards,

2:56:57.260,2:56:59.570
so I can put that in a timeline.

2:56:59.570,2:57:00.600
Et cetera.

2:57:00.600,2:57:03.050
So that's just a super[br]quick demonstration

2:57:03.050,2:57:06.620
of that tool, the resonator.

2:57:06.620,2:57:10.310
Links are all here[br]in the slides.

2:57:10.310,2:57:13.390
And the final tool I wanted[br]to mention very quickly

2:57:13.390,2:57:16.220
is the mix and match tool.

2:57:16.220,2:57:21.980
You remember my explanation[br]about Wikidata as Nexus,

2:57:21.980,2:57:27.320
as connection point between many[br]databases, many data sources.

2:57:27.320,2:57:31.080
Those depend on[br]these equivalencies.

2:57:31.080,2:57:35.300
On Wikidata being taught[br]that this item is like that

2:57:35.300,2:57:37.940
ID in this other database.

2:57:37.940,2:57:41.810
And mix and match is a tool[br]again by, Magnus Manske.

2:57:41.810,2:57:44.690
Maybe you're detecting[br]a pattern here.

2:57:44.690,2:57:47.390
It's a tool by Magnus[br]that is designed

2:57:47.390,2:57:50.270
to enable us to kind[br]of take a foreign,

2:57:50.270,2:57:54.950
an external data set, put[br]it alongside Wikidata,

2:57:54.950,2:57:56.690
and kind of try and align them.

2:57:56.690,2:57:59.410
So this item in this[br]external dataset,

2:57:59.410,2:58:01.230
is that already[br]covered in Wikidata?

2:58:01.230,2:58:02.890
If so, by what queue number?

2:58:02.890,2:58:03.890
By what item?

2:58:03.890,2:58:06.170
If not, maybe we need[br]to create a Wikidata

2:58:06.170,2:58:07.610
item to represent it.

2:58:07.610,2:58:10.010
Or maybe it's a[br]duplicate, or something.

2:58:10.010,2:58:15.980
So the mix and match tool has[br]a list of external data sets,

2:58:15.980,2:58:18.140
as you can see.

2:58:18.140,2:58:21.260
The Art and Architecture[br]Thesaurus by the Getty Research

2:58:21.260,2:58:22.220
Institute.

2:58:22.220,2:58:26.690
Or the Australian[br]Dictionary of Biography.

2:58:26.690,2:58:28.880
All kinds of external[br]data sets here.

2:58:32.470,2:58:40.060
Somewhere here I had a specific[br]link to the Royal Society.

2:58:40.060,2:58:41.710
It can also give[br]me some statistics.

2:58:41.710,2:58:47.410
So there is an external data set[br]of all the Fellows of the Royal

2:58:47.410,2:58:48.001
Society.

2:58:48.001,2:58:48.500
Right?

2:58:48.500,2:58:54.970
The oldest academic[br]learned society in England.

2:58:54.970,2:58:57.415
And the internet is tired.

2:59:03.240,2:59:04.640
Here we go.

2:59:04.640,2:59:07.115
Nope.

2:59:07.115,2:59:08.105
Did that work?

2:59:12.560,2:59:15.390
Fellows of the Royal[br]Society, here we go.

2:59:15.390,2:59:17.970
So this one is complete.

2:59:17.970,2:59:21.330
I mean, people have manually[br]gone over every single item

2:59:21.330,2:59:24.330
there and either[br]matched it to Wikidata

2:59:24.330,2:59:27.390
or declared that it was not[br]in scope, or a duplicate

2:59:27.390,2:59:28.520
or whatever.

2:59:28.520,2:59:31.230
But let's look at site stats.

2:59:31.230,2:59:35.210
This is a fun kind of[br]aspect of this tool.

2:59:35.210,2:59:38.530
But that is not working.

2:59:38.530,2:59:40.820
Or it's taking too long.

2:59:40.820,2:59:43.940
So let's just demonstrate[br]how this works.

2:59:43.940,2:59:45.590
Maybe Britannica?

2:59:45.590,2:59:46.780
Is that done already?

2:59:52.570,2:59:53.990
Here we go.

2:59:53.990,2:59:55.330
Encyclopedia Britannica.

2:59:55.330,2:59:55.960
Yeah.

2:59:55.960,3:00:02.040
So the Encyclopedia[br]Britannica has

3:00:02.040,3:00:05.940
40% of the items there[br]are not yet processed.

3:00:05.940,3:00:07.830
So let's process one of them.

3:00:07.830,3:00:16.180
For example there is an item[br]in the Encyclopedia Britannica

3:00:16.180,3:00:19.960
called Boston, England.

3:00:19.960,3:00:23.050
As you know[br]All-American place names

3:00:23.050,3:00:26.050
are totally stolen[br]from elsewhere.

3:00:26.050,3:00:29.440
So there is a Boston[br]in England, though it's

3:00:29.440,3:00:30.700
no longer the famous one.

3:00:30.700,3:00:36.340
And the mix and match[br]tool has automatically

3:00:36.340,3:00:39.610
matched it based on[br]the label to queue

3:00:39.610,3:00:43.900
100, which is Boston big[br]city in the United States.

3:00:43.900,3:00:45.500
And that is incorrect, right?

3:00:45.500,3:00:48.910
That's kind of naive computer[br]going, well this is Boston,

3:00:48.910,3:00:50.820
and this other thing[br]is also Boston.

3:00:50.820,3:00:56.260
And it is asking me to[br]confirm this match or not.

3:00:56.260,3:00:57.400
You see?

3:00:57.400,3:01:01.120
So this is the Boston,[br]England from Britannica.

3:01:01.120,3:01:04.720
And the tool is asking[br]me, is this the same as

3:01:04.720,3:01:06.910
Boston queue 100 in America?

3:01:06.910,3:01:07.990
The answer is no.

3:01:07.990,3:01:10.110
I removed this.

3:01:10.110,3:01:11.860
I remove this match.

3:01:11.860,3:01:15.430
And now this Boston,[br]England is unmatched.

3:01:15.430,3:01:23.230
And I can match it to the[br]correct one in England.

3:01:23.230,3:01:27.370
I can do this by searching[br]English Wikipedia,

3:01:27.370,3:01:28.780
or searching Wikidata.

3:01:28.780,3:01:32.000
I mean, it has[br]these handy links.

3:01:32.000,3:01:36.910
So the English town[br]is in Lincolnshire.

3:01:36.910,3:01:38.230
Boston, Lincolnshire.

3:01:38.230,3:01:46.030
So I can go there and then[br]get the Wikidata item number.

3:01:46.030,3:01:49.810
See this is not queue[br]100, Boston in the states,

3:01:49.810,3:01:53.440
this is queue 311975[br]town in Lincolnshire.

3:01:53.440,3:01:57.310
I can get this queue[br]number, go back to the mix

3:01:57.310,3:01:58.160
and match tool--

3:01:58.160,3:01:59.110
Where was that?

3:01:59.110,3:02:00.180
Here we are.

3:02:00.180,3:02:01.510
And set queue.

3:02:01.510,3:02:08.650
I can tell the tool that this is[br]the right Boston, and click OK.

3:02:08.650,3:02:14.550
And now this town[br]in Lincolnshire,

3:02:14.550,3:02:17.100
you can see this here,[br]this item, queue 311975,

3:02:17.100,3:02:21.190
is linked to Britannica.

3:02:21.190,3:02:22.660
What does this mean?

3:02:22.660,3:02:23.820
Well, if we go there.

3:02:23.820,3:02:25.380
If we actually go[br]to the Wikidata

3:02:25.380,3:02:28.890
entity you will see[br]that in addition

3:02:28.890,3:02:34.140
to the few statements that[br]it already had, it now has,

3:02:34.140,3:02:38.610
thanks to my clicking, it now[br]has another identifier here.

3:02:38.610,3:02:39.270
See?

3:02:39.270,3:02:43.950
Encyclopedia Britannica[br]Online ID, with this link.

3:02:43.950,3:02:49.440
And if we click it, we[br]will indeed reach this page

3:02:49.440,3:02:51.510
in the Britannica[br]online, which is indeed

3:02:51.510,3:02:53.700
about this town in Lincolnshire.

3:02:53.700,3:02:54.510
You see?

3:02:54.510,3:02:58.650
So I've contributed one[br]of those mappings, one

3:02:58.650,3:03:01.950
of those identifiers,[br]into Wikidata.

3:03:01.950,3:03:04.860
And I didn't have[br]to do it manually.

3:03:04.860,3:03:07.980
This tool kind of prompted[br]me to either confirm

3:03:07.980,3:03:09.480
if it was correct,[br]I could have just

3:03:09.480,3:03:12.150
clicked confirm since[br]it wasn't correct.

3:03:12.150,3:03:16.920
I corrected it manually, but[br]it made this edit on my behalf.

3:03:16.920,3:03:21.180
So that's another tool that[br]encourages us to systematically

3:03:21.180,3:03:24.360
teach Wikidata more things.

3:03:24.360,3:03:25.860
And we're out of time.

3:03:25.860,3:03:29.430
Go edit Wikidata, Now[br]that you have the power,

3:03:29.430,3:03:30.510
you know the deal.

3:03:30.510,3:03:32.430
Use it for good,[br]and not for evil.

3:03:32.430,3:03:35.640
If you have questions,[br]this is my email address.

3:03:35.640,3:03:38.640
If you're watching this video[br]not live the description

3:03:38.640,3:03:41.610
will have links to the[br]slides, and to a bunch

3:03:41.610,3:03:44.610
of other useful[br]pieces of information.

3:03:44.610,3:03:49.510
Any last questions on IRC?

3:03:49.510,3:03:53.210
If not, thank you[br]for your attention.

3:03:53.210,3:03:56.470
And if you like this, and if you[br]feel that you now get Wikidata,

3:03:56.470,3:03:58.330
and you get what it's[br]good for, and you're

3:03:58.330,3:04:01.660
inspired to contribute, I have[br]only one request from you.

3:04:01.660,3:04:04.960
I mean, in addition to using[br]it for good not for evil,

3:04:04.960,3:04:07.630
I ask that you spread the word.

3:04:07.630,3:04:09.550
Show this video--[br]share this video

3:04:09.550,3:04:13.180
with other people in your[br]community, or around you.

3:04:13.180,3:04:16.000
Teach this yourself[br]once you're comfortable

3:04:16.000,3:04:17.650
with these concepts.

3:04:17.650,3:04:21.330
Feel free to use my slides.

3:04:21.330,3:04:23.580
Yeah, and edit Wikidata.

3:04:23.580,3:04:27.010
Thank you very[br]much, and goodbye.