0:00:05.973,0:00:07.908
Hi, guys! Can everybody hear me?
0:00:09.170,0:00:11.898
So, hi! Nice to meet you all.[br]I'm Erica Azzellini.
0:00:11.898,0:00:14.606
I'm one of the Wikimovement [br]Brazil's Liaison,
0:00:14.606,0:00:17.829
and this is my first international [br]Wikimedia event,
0:00:17.829,0:00:21.023
so I'm super excited to be here[br]and I hopefully,
0:00:21.023,0:00:24.311
will share something interesting for you[br]all here on this lengthy talk.
0:00:25.247,0:00:30.441
So this work starts with research [br]that I was developing in Brazil,
0:00:30.441,0:00:34.219
Computational Journalism [br]and Structured Narratives with Wikidata.
0:00:34.276,0:00:35.958
So in journalism,
0:00:35.958,0:00:39.616
they're using some natural language[br]generation software
0:00:39.616,0:00:41.418
for automating news
0:00:41.418,0:00:46.535
for news that have [br]quite similar narrative structure.
0:00:46.535,0:00:51.600
And we developed this concept here [br]of structured narratives,
0:00:51.600,0:00:54.548
thinking about this practice [br]on computational journalism,
0:00:54.548,0:00:58.361
that is the development of verbal text,[br]understandable by humans,
0:00:58.361,0:01:01.274
automated from predetermined [br]arrangements that process information
0:01:01.274,0:01:05.395
from structured databases, [br]which looks like that,
0:01:05.395,0:01:10.043
the Wikimedia universe[br]and on this tool that we developed.
0:01:10.043,0:01:13.555
So, when I'm talking about verbal text[br]understandable by humans,
0:01:13.555,0:01:15.808
I'm talking about Wikipedia entries.
0:01:15.808,0:01:17.778
When I'm talking about [br]structured databases,
0:01:17.778,0:01:20.017
of course, I'm talking about [br]Wikidata here.
0:01:20.017,0:01:22.777
And predetermined arrangement,[br]I'm talking about Mbabel,
0:01:22.777,0:01:24.271
that is this tool.
0:01:25.467,0:01:31.216
The Mbabel tool was inspired by a template[br]by user Pharos, right here in front of me,
0:01:31.279,0:01:33.356
thank you very much,
0:01:33.356,0:01:39.114
and it was developed with Ederporto[br]that is right here too,
0:01:39.114,0:01:40.974
the brilliant Ederporto.
0:01:42.599,0:01:44.498
We developed this tool
0:01:44.498,0:01:47.780
that automatically generates [br]Wikipedia entries
0:01:47.780,0:01:50.600
based on information from Wikidata.
0:01:53.189,0:01:58.130
We actually do some thematic templates
0:01:58.130,0:02:01.152
that are created on the Wikidata module,
0:02:01.573,0:02:03.716
WikidataIB Module,
0:02:03.716,0:02:07.835
and these templates are pre-determined,[br]generic and editable templates
0:02:07.835,0:02:09.677
for various article themes.
0:02:09.677,0:02:15.411
We realized that many Wikipedia entries[br]had a quite similar structured narrative
0:02:15.411,0:02:18.922
so we could create a tool[br]that automatically generates that
0:02:18.922,0:02:21.598
for many Wikidata items.
0:02:24.207,0:02:28.571
Until now we have templates for museums,[br]works of art, books, films,
0:02:28.571,0:02:31.265
journals, earthquakes, libraries,[br]archives,
0:02:31.265,0:02:34.855
and Brazilian municipal [br]and state elections, and growing.
0:02:34.855,0:02:38.984
So, everybody here is able to contribute[br]and create new templates.
0:02:38.984,0:02:43.508
Each narrative template includes[br]an introduction, Wikidata infobox,
0:02:43.508,0:02:46.158
section suggestions for the users,
0:02:46.158,0:02:50.499
content tables or lists with Listeria,[br]depending on the case,
0:02:50.499,0:02:53.713
references and categories,[br]and of course the sentences,
0:02:53.713,0:02:55.776
that are created [br]with the Wikidata information.
0:02:55.776,0:02:58.642
I'm gonna show you in a sec[br]an example of that.
0:03:00.137,0:03:05.749
It's an integration with Wikipedia,[br]integration with Wikidata,
0:03:05.749,0:03:08.760
so the more properties properly filled[br]on Wikidata,
0:03:08.760,0:03:12.311
the more text entries you'll get[br]on your article stub.
0:03:12.857,0:03:15.623
That's very important to highlight here.
0:03:16.343,0:03:18.969
Structuring this Wikidata [br]can get more complex
0:03:18.969,0:03:22.017
as I'm going to show you [br]on the election projects that we've made.
0:03:22.017,0:03:26.552
So I'm going to let you hear this [br]Wikidata Lab XIV for you
0:03:26.552,0:03:29.471
after this lengthy talk
0:03:29.471,0:03:32.259
that is very brief, [br]so you'll be able to choose
0:03:32.259,0:03:34.554
on the work that we've been doing[br]on structuring Wikidata
0:03:34.554,0:03:36.005
for this purpose too.
0:03:37.272,0:03:39.725
We have this challenge to build [br]a narrative template
0:03:39.725,0:03:44.383
that is generic enough [br]to cover different Wikidata items
0:03:44.383,0:03:46.347
and to suppress the gender
0:03:46.347,0:03:50.359
and the number of difficulties [br]of languages,
0:03:52.054,0:03:54.252
and still sounding natural for the user
0:03:54.252,0:03:59.252
because we don't want to sound like[br]it doesn't click for the user
0:03:59.252,0:04:00.546
to edit after that.
0:04:01.956,0:04:07.625
This is how the Mbabel looks like[br]on the bottom form.
0:04:07.625,0:04:14.507
You just have insert the item number there[br]and call the desired template
0:04:14.507,0:04:21.673
and then you have article to edit[br]and expand, and everything.
0:04:22.135,0:04:26.856
So, more importantly, why we did it?[br]Not because it's cool to develop
0:04:26.856,0:04:30.922
things here in Wikidata,[br]we know, we all hear, know about it.
0:04:30.922,0:04:36.178
But we are experimenting this integration[br]from Wikidata to Wikipedia
0:04:36.178,0:04:39.226
and we want to focus [br]on meaningful individual contributions.
0:04:39.226,0:04:42.608
So we've been working [br]on education programs
0:04:42.608,0:04:45.067
and we want the students to feel the value
0:04:45.067,0:04:47.280
of their entries too, but not only--
0:04:47.280,0:04:49.405
Oh, five minutes only,[br]Geez, I'm gonna rush here.
0:04:49.405,0:04:50.599
(laughing)
0:04:50.794,0:04:54.160
And we want you all to make tasks[br]for users in general,
0:04:54.270,0:04:57.801
especially on tables [br]and this kind of content
0:04:57.801,0:04:59.988
that it's a bit of a rush to do.
0:05:02.456,0:05:05.523
And we're working on this concept[br]of abstract Wikipedia.
0:05:05.523,0:05:09.269
Denny Vrandečić wrote an article[br]super interesting about it
0:05:09.269,0:05:11.500
so I linked here too.
0:05:11.500,0:05:14.792
And we also want to now support [br]small language communities
0:05:14.792,0:05:17.845
to fill the lack of content there.
0:05:18.784,0:05:23.885
This is an example of how we've been using[br]this Mbabel tool for GLAM
0:05:23.885,0:05:25.748
and education programs,
0:05:25.748,0:05:29.861
and I showed you earlier[br]the bottom form of the Mbabel tool
0:05:29.861,0:05:34.264
but also we can make red links[br]that aren't exactly empty.
0:05:34.264,0:05:35.931
So you click on this red link
0:05:35.931,0:05:38.862
and you automatically have [br]this article draft
0:05:38.862,0:05:41.660
on your user page to edit.
0:05:42.964,0:05:48.762
And I'm going to briefly talk about it[br]because I only have some minutes more.
0:05:50.009,0:05:51.356
On educational projects,
0:05:51.356,0:05:56.799
we've been doing this with elections [br]in Brazil for journalism students.
0:05:56.799,0:06:01.993
We have the experience[br]with the [inaudible] students
0:06:02.087,0:06:05.314
with user Joalpe--[br]he's not here right now,
0:06:05.314,0:06:07.867
but we all know him, I think.
0:06:07.867,0:06:11.930
And we realize that we have the data[br]about Brazilian elections
0:06:11.930,0:06:14.748
but we don't have media cover on it.
0:06:15.049,0:06:18.249
So we were lacking also [br]Wikipedia entries on it.
0:06:19.029,0:06:23.000
How do we insert this meaningful [br]information on Wikipedia
0:06:23.000,0:06:24.672
that people really access?
0:06:24.672,0:06:27.989
Next year we're going [br]to have some election,
0:06:27.989,0:06:30.710
people are going to look for [br]this kind of information on Wikipedia
0:06:30.710,0:06:32.433
and they simply won't find it.
0:06:32.433,0:06:35.726
So this tool looks quite useful[br]for this purpose
0:06:35.726,0:06:40.214
and the students were introduced,[br]not only to Wikipedia,
0:06:40.214,0:06:42.701
but also to Wikidata.
0:06:42.701,0:06:46.575
Actually, they were introduced [br]to Wikipedia with Wikidata,
0:06:46.575,0:06:50.675
which is an experience super interesting[br]and we had a lot of fun,
0:06:50.675,0:06:52.823
and it was quite challenging [br]to organize all that.
0:06:52.823,0:06:54.513
We can talk about it later too.
0:06:54.979,0:06:58.582
And they also added the background [br]and the analysis sections
0:06:58.582,0:07:01.663
on these elections articles,
0:07:01.663,0:07:05.336
because we don't want them[br]to just simply automate the content there.
0:07:05.336,0:07:06.660
We can do better.
0:07:06.660,0:07:09.247
So this is the example [br]I'm going to show you.
0:07:09.247,0:07:13.106
This is from a municipal election[br]in Brazil.
0:07:15.603,0:07:17.121
Two minutes... oh my!
0:07:18.577,0:07:23.268
This example here was entirely created[br]with the Mbabel tool.
0:07:23.268,0:07:29.496
You have here this introduction text.[br]It really sounds natural for the reader.
0:07:29.496,0:07:32.165
The Wikidata infobox here--
0:07:32.165,0:07:34.907
it's a masterpiece [br]of Ederporto right there.
0:07:34.907,0:07:36.769
(laughter)
0:07:37.438,0:07:42.456
And we have here the tables with the[br]election results for each position.
0:07:42.456,0:07:46.415
And we also have these results here [br]on the textual form too,
0:07:46.415,0:07:51.767
so it really looks like an article[br]that was made, that was handcrafted.
0:07:53.893,0:07:57.814
The references here were also made [br]with the Mbabel tool
0:07:57.814,0:08:01.393
and we used identifiers[br]to build these references here
0:08:01.393,0:08:03.167
and the categories too.
0:08:10.726,0:08:14.999
So, to wrap things up here,[br]it is still a work in progress,
0:08:14.999,0:08:19.326
and we have some challenges [br]on outreach and technical
0:08:19.326,0:08:22.999
to bring Mbabel [br]to other language communities,
0:08:22.999,0:08:24.844
especially the smaller ones,
0:08:24.844,0:08:27.210
and how do we support those tools
0:08:27.210,0:08:29.819
on lower resource [br]language communities too.
0:08:29.819,0:08:33.991
And finally, is it possible [br]to create an Mbabel
0:08:33.991,0:08:36.261
that overcomes language barriers?
0:08:36.261,0:08:39.740
I think that's a question [br]very interesting for the conference
0:08:39.740,0:08:43.835
and hopefully we can figure [br]that out together.
0:08:44.818,0:08:49.799
So, thank you very much,[br]and look for the Mbabel poster downstairs
0:08:49.799,0:08:53.615
if you like to have all this information[br]wrapped up, okay?
0:08:53.615,0:08:55.038
Thank you.
0:08:55.288,0:08:57.564
(audience clapping)
0:09:00.311,0:09:02.778
(moderator) I'm afraid [br]we're a little too short for questions
0:09:02.778,0:09:05.783
but yes, Erica, as she said, [br]has a poster and is very friendly.
0:09:05.783,0:09:07.518
So I'm sure you can talk to her[br]afterwards,
0:09:07.518,0:09:09.389
and if there's time at the end, [br]I'll allow it.
0:09:09.389,0:09:12.131
But in the meantime, [br]I'd like to bring up our next speaker...
0:09:12.237,0:09:13.611
Thank you.
0:09:15.549,0:09:17.140
(audience chattering)
0:09:23.058,0:09:27.016
Next we've got Yolanda Gil, [br]talking about Wikidata and Geosciences.
0:09:27.908,0:09:29.031
Thank you.
0:09:29.031,0:09:31.624
I come from the University [br]of Southern California
0:09:31.624,0:09:35.164
and I've been working with [br]Semantic Technologies for a long time.
0:09:35.164,0:09:37.894
I want to talk about geosciences[br]in particular,
0:09:37.894,0:09:41.225
where this idea of crowd-sourcing[br]from the community is very important.
0:09:41.791,0:09:45.033
So I'll give you a sense [br]that individual scientists,
0:09:45.033,0:09:47.070
most of them in colleges,
0:09:47.070,0:09:50.085
collect their own data [br]for their particular project.
0:09:50.085,0:09:51.932
They describe it in their own way.
0:09:51.932,0:09:55.352
They use their own properties, [br]their own metadata characteristics.
0:09:55.352,0:09:58.560
This is an example [br]of some collaborators of mine
0:09:58.560,0:10:00.124
that collect data from a river.
0:10:00.124,0:10:02.091
They have their own sensors, [br]their own robots,
0:10:02.091,0:10:05.339
and they study the water quality.
0:10:05.339,0:10:11.423
I'm going to talk today about an effort[br]that we did to crowdsource metadata
0:10:11.423,0:10:14.712
for a community that works[br]in paleoclimate.
0:10:14.712,0:10:17.747
The article just came out[br]so it's in the slides if you're curious,
0:10:17.747,0:10:20.619
but it's a pretty large community[br]that work together
0:10:20.619,0:10:24.042
to integrate data more efficiently[br]through crowdsourcing.
0:10:24.042,0:10:28.631
So, if you've heard of the [br]hockey stick graphics for climate,
0:10:28.631,0:10:31.680
this is the community that does this.
0:10:31.680,0:10:34.520
This is a study for climate[br]in the last 200 years,
0:10:34.520,0:10:38.188
and it takes them literally many years[br]to look at data
0:10:38.188,0:10:39.618
from different parts of the globe.
0:10:39.618,0:10:42.607
Each dataset is collected by [br]a different investigator.
0:10:42.699,0:10:44.433
The data is very, very different,
0:10:44.433,0:10:47.017
so it takes them a long time [br]to put together
0:10:47.017,0:10:49.230
these global studies of climate,
0:10:49.230,0:10:51.665
and our goal is to make that [br]more efficient.
0:10:51.665,0:10:53.690
So, I've done a lot of work [br]over the years.
0:10:53.690,0:10:56.585
Going back to 2005, we used to call it,
0:10:56.585,0:10:59.615
"Knowledge Collection from Web Volunteers"
0:10:59.615,0:11:02.236
or from netizens at that time.
0:11:02.236,0:11:04.267
We had a system called "Learner."
0:11:04.267,0:11:07.048
It collected 700,000 common sense,
0:11:07.048,0:11:09.368
common knowledge statements [br]about the world.
0:11:09.368,0:11:11.367
We did a lot of different techniques.
0:11:11.367,0:11:15.333
The forms that we did[br]to extract knowledge from volunteers
0:11:15.333,0:11:19.136
really fit the knowledge models,[br]the data models that we used
0:11:19.136,0:11:21.381
and the properties that we wanted to use.
0:11:21.381,0:11:25.051
I worked with Denny [br]in the system called "Shortipedia"
0:11:25.051,0:11:27.259
when he was a Post Doc at ISI,
0:11:27.259,0:11:31.946
looking at keeping track [br]of the prominence of the assertions,
0:11:31.946,0:11:35.129
and we started to build [br]on Semantic Media Wiki software.
0:11:35.129,0:11:37.113
So everything that [br]I'm going to describe today
0:11:37.113,0:11:38.936
builds on that software,
0:11:38.936,0:11:41.117
but I think that now we have Wikibase,
0:11:41.117,0:11:43.676
we'll be starting to work more [br]on Wikibase.
0:11:43.676,0:11:48.935
So the LinkedEarth is the project[br]where we work with paleoclimate scientists
0:11:48.935,0:11:50.636
to crowdsource the metadata,
0:11:50.636,0:11:54.328
and seeing the title that we said,[br]"controlled crowdsourcing."
0:11:54.328,0:11:57.101
So we found a nice niche
0:11:57.101,0:12:00.538
where we could let them create [br]new properties
0:12:00.538,0:12:02.599
but we had an editorial process for it.
0:12:02.599,0:12:04.444
So I'll describe to you how it works.
0:12:04.444,0:12:10.055
For them, if you're looking at a sample[br]from lake sediments from 200 years ago,
0:12:10.055,0:12:12.622
you use different properties[br]to describe it
0:12:12.622,0:12:15.692
than if you have coral sediments[br]that you're looking at
0:12:15.692,0:12:18.979
or coral samples that you're looking at[br]that you extract from the ocean.
0:12:18.979,0:12:23.532
Palmyra is a coral atoll in the Pacific.
0:12:23.532,0:12:27.918
So if you have coral, you care [br]about the species and the genus,
0:12:27.918,0:12:31.691
but if you're just looking at lake sand,[br]you don't have that.
0:12:31.691,0:12:35.313
So each type of sample [br]has very different properties.
0:12:35.313,0:12:38.798
In LinkedEarth, [br]they're able to see in a map
0:12:38.798,0:12:40.264
where the datasets are.
0:12:40.264,0:12:45.500
They actually annotate their own datasets[br]or the datasets of other researchers
0:12:45.500,0:12:46.787
when they're using it.
0:12:46.787,0:12:50.254
So they have a reason [br]why they want certain properties
0:12:50.254,0:12:52.289
to describe those datasets.
0:12:52.289,0:12:56.683
Whenever there are disagreements, [br]or whenever there are agreements,
0:12:56.683,0:12:58.595
there's community discussions [br]about them
0:12:58.595,0:13:02.894
and they're also polls to decide on [br]what properties to settle.
0:13:02.894,0:13:05.659
So it's a nice ecosystem. [br]I'll give you examples.
0:13:05.659,0:13:11.322
You look at a particular dataset,[br]in this case it's a lake in Africa.
0:13:11.322,0:13:14.241
So you have the category of the page;[br]it can be a dataset,
0:13:14.241,0:13:15.491
it can be other things.
0:13:15.491,0:13:21.181
You can download the dataset itself[br]and you have kind of canonical properties
0:13:21.181,0:13:23.737
that they have all agreed to have [br]for datasets,
0:13:23.737,0:13:25.992
and then under Extra Information,
0:13:25.992,0:13:29.369
those are properties [br]that the person describing this dataset,
0:13:29.369,0:13:31.007
added on their own accord.
0:13:31.007,0:13:32.628
So these can be new properties.
0:13:32.628,0:13:36.730
We call them "crowd properties,"[br]rather than "core properties."
0:13:37.291,0:13:41.319
And then when you're describing [br]your dataset,
0:13:41.319,0:13:43.774
in this case [br]it's an ice core that you got
0:13:43.774,0:13:45.716
from a glacier dataset,
0:13:45.765,0:13:49.178
and your'e adding a dataset [br]you want to talk about measurements,
0:13:49.178,0:13:54.073
you have an offering [br]of all the existing properties
0:13:54.073,0:13:55.278
that match what you're saying.
0:13:55.278,0:13:58.409
So we do this search completion[br]so that you can adopt that.
0:13:58.409,0:14:00.140
That promotes normalization.
0:14:00.140,0:14:04.260
The core of the properties[br]has been agreed by the community
0:14:04.260,0:14:06.220
so we're really extending that core.
0:14:06.220,0:14:08.795
And that core is very important[br]because it gives structure
0:14:08.795,0:14:10.735
to all the extensions.
0:14:10.735,0:14:14.382
We engage the community [br]through many different ways.
0:14:14.382,0:14:17.260
We had one face-to-face meeting[br]at the beginning
0:14:17.260,0:14:21.611
and after about a year and a half,[br]we do have a new standard,
0:14:21.611,0:14:25.154
and a new way for them[br]to continue to evolve that standard.
0:14:25.154,0:14:30.569
They have editors, very much[br]in the Wikipedia style
0:14:30.569,0:14:31.582
of editorial boards.
0:14:31.582,0:14:34.098
They have working groups [br]for different types of data.
0:14:34.098,0:14:36.090
They do polls with the community,
0:14:36.090,0:14:40.879
and they have pretty nice engagement[br]of the community at large,
0:14:40.879,0:14:43.706
even if they've never visited our Wiki.
0:14:43.706,0:14:46.183
The metadata evolves
0:14:46.183,0:14:48.775
so what we do is that people annotate[br]their datasets,
0:14:48.775,0:14:52.321
then the schema evolves,[br]the properties evolve
0:14:52.321,0:14:55.379
and we have an entire infrastructure[br]and mechanisms
0:14:55.379,0:15:00.336
to re-annotate the datasets[br]with the new structure of the ontology
0:15:00.336,0:15:01.711
and the new properties.
0:15:01.711,0:15:05.210
This is described in the paper.[br]I won't go into the details.
0:15:05.210,0:15:07.583
But I think that [br]having that kind of capability
0:15:07.583,0:15:10.342
in Wikibase would be really interesting.
0:15:10.342,0:15:14.041
We basically extended [br]Semantic Media Wiki and Media Wiki
0:15:14.041,0:15:15.722
to create our own infrastructure.
0:15:15.722,0:15:18.855
I think a lot of this is now something[br]that we find in Wikibase,
0:15:18.961,0:15:20.615
but this is older than that.
0:15:20.615,0:15:24.999
And in general, we have many projects[br]where we look at crowdsourcing
0:15:24.999,0:15:29.885
not just descriptions of datasets[br]but also descriptions of hydrology models,
0:15:29.885,0:15:33.563
descriptions of multi-step [br]data analytic workflows
0:15:33.563,0:15:36.080
and many other things in the sciences.
0:15:36.080,0:15:42.833
So we are also interested in including[br]in Wikidata additional things
0:15:42.833,0:15:46.250
that are not just datasets or entities
0:15:46.250,0:15:48.512
but also other things [br]that have to do with science.
0:15:48.512,0:15:53.770
I think Geosciences are more complex[br]in this sense than Biology, for example.
0:15:54.923,0:15:56.233
That's it.
0:15:56.513,0:15:57.885
Thank you.[br](audience clapping)
0:16:01.640,0:16:03.772
- Do I have time for questions?[br]- Yes.
0:16:03.772,0:16:06.871
(moderator) We have time [br]for just a couple of short questions.
0:16:07.751,0:16:11.342
When answering, [br]can go back to the microphone?
0:16:12.529,0:16:14.520
- Yes.[br]- Hopefully, yeah.
0:16:21.314,0:16:25.002
(audience 1) Does the structure allow[br]tabular datasets to be described
0:16:25.002,0:16:26.988
and can you talk a bit about that?
0:16:27.225,0:16:32.667
Yes. So the properties of the datasets[br]talk more about who collected them,
0:16:32.667,0:16:36.759
what kind of data was collected,[br]what kind of sample it was,
0:16:36.759,0:16:39.790
and then there's a separate standard[br]which is called "lipid"
0:16:39.790,0:16:43.065
that's complementary and mapped[br]to the properties
0:16:43.065,0:16:46.994
that describes the format[br]of the actual files
0:16:47.075,0:16:49.343
and the actual structure of the data.
0:16:49.343,0:16:53.631
So, you're right that there's both,[br]"how do I find data about x"
0:16:53.631,0:16:55.557
but also, "Now, how do I use it?
0:16:55.557,0:17:00.211
How do I know where[br]the temperature that I'm looking for
0:17:00.211,0:17:03.013
is actually in the file?"
0:17:03.656,0:17:05.394
(moderator) This will be the last.
0:17:06.887,0:17:09.034
(audience 2) I'll have [br]to make it relevant.
0:17:09.504,0:17:15.667
So, you have shown this process [br]of how users can suggest
0:17:15.667,0:17:18.985
or like actually already put in [br]properties,
0:17:18.985,0:17:22.705
and I didn't fully understand[br]how this thing works,
0:17:22.705,0:17:24.027
or what's the process behind it.
0:17:24.027,0:17:28.045
Is there some kind of [br]folksonomy approach--obviously--
0:17:28.045,0:17:33.387
but how is it promoted [br]into the core vocabulary
0:17:33.387,0:17:36.255
if something is promoted?
0:17:36.255,0:17:37.882
Yes, yes. It is.
0:17:37.882,0:17:42.202
So what we do is we have a core ontology[br]and the initial one was actually
0:17:42.202,0:17:45.618
very thoughtfully put together [br]through a lot of discussion
0:17:45.618,0:17:47.964
by very few people.
0:17:47.964,0:17:51.052
And then the idea was [br]the whole community can extend that
0:17:51.052,0:17:52.971
or propose changes to that.
0:17:52.971,0:17:56.919
So, as they are describing datasets,[br]they can add new properties
0:17:56.919,0:17:59.526
and those become "crowd properties."
0:17:59.526,0:18:02.941
And every now and then, [br]the Editorial Committee
0:18:02.941,0:18:04.367
looks at all of those properties,
0:18:04.367,0:18:07.795
the working groups look at all of those[br]crowd properties,
0:18:07.795,0:18:11.714
and decide whether to incorporate them[br]into the main ontology.
0:18:11.714,0:18:15.804
So it could be because they're used[br]for a lot of dataset descriptions.
0:18:15.804,0:18:18.920
It could be because [br]they are proposed by somebody
0:18:18.920,0:18:23.339
and they're found to be really interesting[br]or key, or uncontroversial.
0:18:23.339,0:18:30.267
So there's an entire editorial process[br]to incorporate those new crowd properties
0:18:30.267,0:18:32.188
or the folksonomy part of it,
0:18:32.188,0:18:36.308
but they are really built around the core[br]of the ontology.
0:18:36.404,0:18:40.280
The core ontology then grows[br]with more crowd properties
0:18:40.280,0:18:44.311
and then people propose[br]additional crowd properties again.
0:18:44.311,0:18:46.979
So we've gone through a couple [br]of these iterations
0:18:46.979,0:18:51.386
of rolling out a new core,[br]and then extending it,
0:18:51.386,0:18:55.570
and then rolling out a new core[br]and then extending it.
0:18:55.570,0:18:57.779
- (audience 2) Great. Thank you.[br]- Thanks.
0:18:57.779,0:19:00.437
(moderator) Thank you.[br](audience applauding)
0:19:02.295,0:19:03.777
(moderator) Thank you, Yolanda.
0:19:03.777,0:19:07.494
And now we have Adam Shorn[br]with "Something About Wikibase,"
0:19:07.599,0:19:09.299
according to the title.
0:19:09.708,0:19:12.956
Uh... where's the internet? There it is.
0:19:13.245,0:19:18.925
So, I'm going to do a live demo,[br]which is probably a bad idea
0:19:18.925,0:19:21.362
but I'm going to try and do it[br]as the birthday present later
0:19:21.362,0:19:24.268
so I figure I might as well try it here.
0:19:24.292,0:19:27.304
And I also have some notes on my phone[br]because I have no slides.
0:19:29.349,0:19:32.248
So, two years ago, [br]I made these Wikibase doc images
0:19:32.248,0:19:34.052
that quite a few people have tried out,
0:19:34.052,0:19:38.087
and even before then, [br]I was working on another project,
0:19:38.087,0:19:42.363
which is kind of ready now, [br]and here it is.
0:19:43.690,0:19:46.832
It's a website that allows you [br]to instantly create a Wikibase
0:19:46.900,0:19:48.930
with a query service and quick statements,
0:19:48.930,0:19:51.616
without needing to know about[br]any of the technical details,
0:19:51.616,0:19:54.295
without needing to manage [br]any of them either.
0:19:54.295,0:19:57.054
There are still lots of features to go[br]and there's still some bugs,
0:19:57.054,0:19:59.348
but here goes the demo.
0:19:59.348,0:20:02.628
Let me get my emails up ready...[br]because I need them too...
0:20:03.315,0:20:06.514
Da da da... Stopwatch.
0:20:07.272,0:20:08.488
Okay.
0:20:08.829,0:20:14.253
So it's a simple as... [br]at the moment it's locked down behind...
0:20:14.337,0:20:16.495
Oh no! German keyboard!
0:20:16.495,0:20:18.703
(audience laughing)
0:20:22.556,0:20:23.923
Foiled... okay.
0:20:24.955,0:20:26.214
Okay.
0:20:26.634,0:20:28.417
(audience continues to laugh)
0:20:30.434,0:20:31.989
Aha! Okay.
0:20:32.950,0:20:35.335
I'll remember that for later.[br](laughs)
0:20:36.911,0:20:38.119
Yes.
0:20:39.438,0:20:40.855
♪ (humming) ♪
0:20:40.961,0:20:44.932
Oh my god... now it's American.
0:20:53.871,0:20:56.131
All you have to do is create an account...
0:20:58.570,0:21:00.007
da da da...
0:21:00.566,0:21:02.432
Click this button up here...
0:21:02.478,0:21:05.512
Come up with a name for Wiki--[br]"Demo1"
0:21:05.862,0:21:07.299
"Demo1"
0:21:07.568,0:21:09.135
"Demo user"
0:21:09.203,0:21:11.864
Agree to the terms [br]which don't really exist yet.
0:21:12.298,0:21:14.247
(audience laughing)
0:21:15.264,0:21:17.698
Click on this thing which isn't a link.
0:21:21.519,0:21:23.886
And then you have your Wikibase.
0:21:23.886,0:21:26.602
(audience cheers and claps)
0:21:28.554,0:21:30.421
Anmelden in German.
0:21:30.421,0:21:35.126
Demo... oh god! I'm learning lots about[br]my demo later.
0:21:35.569,0:21:40.069
1-6-1-4-S-G...
0:21:40.166,0:21:42.567
- (audience 3) Y...[br]- (Adam) It's random.
0:21:43.016,0:21:44.567
(audience laughing)
0:21:46.237,0:21:47.958
Oh, come on....[br](audience laughing)
0:21:48.001,0:21:50.543
Oh no. It's because this is a capital U...
0:21:51.333,0:21:53.283
(audience chattering)
0:21:54.453,0:21:56.545
6-1-4....
0:21:57.465,0:22:01.248
S-G-ENJ...
0:22:01.623,0:22:03.794
Is J... oh no. That's... oh yeah. Okay.
0:22:03.843,0:22:06.242
I'm really... I'm gonna have to look[br]at the laptop
0:22:06.242,0:22:07.836
that I'm doing this on later.
0:22:07.836,0:22:09.129
Cool...
0:22:11.046,0:22:13.709
Da da da da da...
0:22:14.687,0:22:17.040
Maybe I should have some things[br]in my clipboard ready.
0:22:17.539,0:22:19.093
Okay, so now I'm logged in.
0:22:22.631,0:22:25.065
Oh... keyboards.
0:22:28.083,0:22:30.012
So you can go and create an item...
0:22:36.194,0:22:38.508
Yeah, maybe I should make a video.[br]It might be easier.
0:22:38.927,0:22:42.207
So, yeah. You can make items,[br]you have quick statements here
0:22:42.207,0:22:43.901
that have... oh... it is all in German.
0:22:43.901,0:22:45.088
(audience laughing)
0:22:45.088,0:22:46.297
(sighs)
0:22:46.926,0:22:49.021
Oh, log in? Log in?
0:22:50.348,0:22:52.088
It has... Oh, set up ready.
0:22:52.088,0:22:53.482
Da da da...
0:22:55.965,0:22:57.850
It's as easy as...
0:22:58.966,0:23:01.350
I learned how to use [br]Quick Statements yesterday...
0:23:01.350,0:23:03.245
that's what I know how to do.
0:23:04.657,0:23:07.089
I can then go back to the Wiki...
0:23:08.008,0:23:09.804
We can go and see in Recent Changes
0:23:09.804,0:23:11.942
that there are now two items,[br]the one that I made
0:23:11.942,0:23:13.759
and the one from Quick Statements...
0:23:13.759,0:23:14.881
and then you go to Quick...
0:23:14.881,0:23:16.511
♪ (hums a tune) ♪
0:23:17.637,0:23:18.770
Stop...no...
0:23:18.927,0:23:20.120
No...
0:23:20.454,0:23:22.437
(audience laughing)
0:23:28.394,0:23:30.006
Oh god...
0:23:30.061,0:23:32.012
I'm glad I tried this out in advance.
0:23:33.464,0:23:35.678
There you go. [br]And the query service is updated.
0:23:35.830,0:23:37.763
(audience clapping)
0:23:42.357,0:23:45.359
And the idea of this is it'll allow [br]people to try out Wikibases.
0:23:45.359,0:23:48.493
Hopefully, it'll even be able [br]to allow people to...
0:23:49.110,0:23:50.945
have their real Wikibases here.
0:23:50.945,0:23:53.783
At the moment you can create[br]as many as you want
0:23:53.783,0:23:55.653
and they all just appear [br]in this lovely list.
0:23:55.653,0:23:59.182
As I said, there's lots of bugs[br]but it's all super quick.
0:23:59.914,0:24:03.392
Exactly how this is going to continue[br]in the future, we don't know yet
0:24:03.392,0:24:05.757
because I only finished writing this[br]in the last few days.
0:24:05.757,0:24:09.286
It's currently behind an invitation code[br]so that if you want to come try it out,
0:24:09.286,0:24:10.888
come and talk to me.
0:24:11.645,0:24:15.730
And if you have any other comments[br]or thoughts, let me know.
0:24:15.861,0:24:19.711
Oh, three minutes...40. That's...[br]That's not that bad.
0:24:19.986,0:24:21.022
Thanks.
0:24:21.022,0:24:22.622
(audience clapping)
0:24:28.435,0:24:30.006
Any questions?
0:24:31.020,0:24:35.553
(audience 5) Does the Quick Statements[br]and the Query Service
0:24:35.553,0:24:38.602
are automatically updated?
0:24:39.553,0:24:42.345
Yes. So the idea is that [br]there will be somebody,
0:24:42.345,0:24:43.500
at the moment, me,
0:24:43.500,0:24:45.144
maintaining all of the horrible stuff
0:24:45.144,0:24:47.290
that you don't have to behind the scenes.
0:24:47.657,0:24:50.157
So kind of think of it like GitHub.com,
0:24:50.157,0:24:54.058
but you don't have to know anything [br]about Git to use it. It's just all there.
0:24:55.241,0:24:56.886
- [inaudible][br]- Yeah, we'll get that.
0:24:56.886,0:25:00.247
But any of those [br]big hosted solution things.
0:25:00.833,0:25:03.263
- (audience 6) A feature request.[br]- Yes.
0:25:03.263,0:25:05.479
Is there any-- In Scope
0:25:05.479,0:25:09.799
do you have plans on making it[br]so you can easily import existing...
0:25:09.799,0:25:12.549
- Wikidata...[br]- I have loads of plans.
0:25:12.549,0:25:14.909
Like I want there to be a button[br]where you can just import
0:25:14.909,0:25:17.348
another whole Wikibase and all of--yeah.
0:25:17.436,0:25:20.723
There will, in the future list [br]that's really long. Yeah.
0:25:24.454,0:25:28.406
(audience 7) I understand that it's...[br]you want to make it user-friendly
0:25:28.406,0:25:32.242
but if I want to access [br]to the machine itself, can I do that?
0:25:32.242,0:25:34.673
Nope.[br](audience laughing)
0:25:37.006,0:25:40.863
So again, like, in the longer term future,[br]there are possib...
0:25:40.863,0:25:43.810
Everything's possible, [br]but at the moment, no.
0:25:45.156,0:25:49.743
(audience 8) Two questions. [br]Is there a plan to have export tools
0:25:49.743,0:25:52.791
so that you can export it [br]to your own Wikibase maybe at some point?
0:25:52.791,0:25:53.824
- Yes.[br]- Great.
0:25:53.824,0:25:55.565
And is this a business?
0:25:56.003,0:25:58.164
I have no idea.[br](audience laughing)
0:26:00.015,0:26:01.545
Not currently.
0:26:05.754,0:26:08.451
(audience 9) What if I stop [br]using it tomorrow,
0:26:08.451,0:26:11.096
how long will the data be there?
0:26:11.181,0:26:14.632
So my plan was at the end of WikidataCon[br]I was going to delete all of the data
0:26:14.632,0:26:18.060
and there's a Wikibase Workshop[br]on a Sunday,
0:26:18.060,0:26:21.671
and we will maybe be using this[br]for the Wikibase workshop
0:26:21.671,0:26:23.801
so that everyone can have[br]their own Wikibase.
0:26:23.801,0:26:27.366
And then, from that point,[br]I probably won't be deleting the data
0:26:27.366,0:26:29.008
so it will all just stay there.
0:26:31.763,0:26:32.923
(moderator) Question.
0:26:34.524,0:26:36.114
(audience 10) It's two minutes...
0:26:36.175,0:26:39.505
Alright, fine. I'll allow two more [br]questions if you talk quickly.
0:26:39.505,0:26:41.550
(audience laughing)
0:26:47.370,0:26:49.999
- Alright, good people.[br]- Thank you, Adam.
0:26:49.999,0:26:52.418
Thank you for letting me test [br]my demo... I mean...
0:26:52.418,0:26:54.640
I'm going to do it different.[br](audience clapping)
0:26:59.512,0:27:00.753
(moderator) Thank you.
0:27:00.753,0:27:03.869
Now we have Dennis Diefenbach [br]presenting Q Answer.
0:27:04.489,0:27:08.129
Hello, I'm Dennis Diefenbach,[br]I would like to present Q-Answer
0:27:08.129,0:27:11.392
which is a question-answering system[br]on top of Wikidata.
0:27:11.392,0:27:16.203
So, what we need are some questions [br]and this is the interface of QAnswer.
0:27:16.203,0:27:23.460
For example, where is WikidataCon?
0:27:23.901,0:27:25.975
Alright, I think it's written like this.
0:27:27.432,0:27:32.432
2019... And we get this response[br]which is Berlin.
0:27:32.458,0:27:38.425
So, other questions. For example,[br]"When did Wikidata start?"
0:27:38.430,0:27:42.383
It started the 30 October 2012[br]so it's birthday is approaching.
0:27:44.079,0:27:48.014
It is 6 years old, [br]so it will be their 7th birthday.
0:27:49.133,0:27:51.583
Who is developing Wikidata?
0:27:51.583,0:27:54.371
The Wikimedia Foundation [br]and Wikimedia Deutschland,
0:27:54.371,0:27:55.988
so thank you very much to them.
0:27:57.013,0:28:02.947
Something like museums in Berlin...[br]I don't know why this is not so...
0:28:05.494,0:28:07.737
Only one museum... no, yeah, a few more.
0:28:09.167,0:28:10.995
So, when you ask something like this,
0:28:10.995,0:28:14.178
we allow the user [br]to explore the information
0:28:14.178,0:28:16.308
with different aggregations.
0:28:16.308,0:28:18.953
For example, [br]if there are many geo coordinates
0:28:18.953,0:28:21.476
attached to the entities,[br]we will display a map.
0:28:21.476,0:28:26.357
If there are many images attached to them,[br]we will display the images,
0:28:26.357,0:28:29.057
and otherwise there is a list[br]where you can explore
0:28:29.057,0:28:30.855
the different entities.
0:28:33.236,0:28:35.605
You can ask something like [br]"Who is the mayor of Berlin,"
0:28:36.643,0:28:40.201
"Give me politicians born in Berlin,"[br]and things like this.
0:28:40.201,0:28:44.428
So you can both ask keyword questions[br]and foreign natural language questions.
0:28:45.171,0:28:48.604
The whole data is coming from Wikidata
0:28:48.604,0:28:55.346
so all entities which are in Wikidata[br]are queryable by this service.
0:28:55.869,0:28:59.244
And the data is really all from Wikidata
0:28:59.244,0:29:01.207
in the sense, [br]there are some Wikipedia snippets,
0:29:01.207,0:29:04.851
there are images from Wikimedia Commons,
0:29:04.851,0:29:07.644
but the rest is all Wikidata data.
0:29:08.760,0:29:11.678
We can do this in several languages.[br]This is now in Chinese.
0:29:11.678,0:29:15.441
I don't know what is written there[br]so do not ask me.
0:29:15.441,0:29:19.893
We are currently supporting this languages[br]with more or less good quality
0:29:19.893,0:29:22.094
because... yeah.
0:29:23.332,0:29:27.563
So, how can this be useful[br]for the Wikidata community?
0:29:27.968,0:29:30.052
I think there are different reasons.
0:29:30.052,0:29:33.786
First of all, this thing helps you[br]to generate SPARQL queries
0:29:33.786,0:29:37.043
and I know there are even some workshops[br]about how to use SPARQL.
0:29:37.043,0:29:39.444
It's not a language that everyone speaks.
0:29:39.444,0:29:45.147
So, if you ask something like [br]"a philosopher born before 1908,"
0:29:45.147,0:29:48.697
to figure out, to construct [br]a SPARQL query like this could be tricky,
0:29:50.001,0:29:54.257
In fact when you ask a question,[br]we generate many SPARQL queries
0:29:54.301,0:29:57.486
and the first one is always the thing,[br]the SPARQL query where we think
0:29:57.486,0:29:59.008
this is the good one.
0:29:59.017,0:30:02.651
So, if you ask your question[br]and then you go on SPARQL list,
0:30:02.691,0:30:06.468
then there is this button [br]for the Wikidata query service
0:30:06.468,0:30:11.811
and you have the SPARQL query right there[br]and you will get the same result
0:30:11.811,0:30:15.184
as you would get in the interface.
0:30:16.906,0:30:19.289
Another thing where it could be useful for
0:30:19.289,0:30:23.468
is for finding missing [br]contextual information.
0:30:23.468,0:30:27.057
For example, if you ask for actors[br]in "The Lord of the Rings,"
0:30:27.057,0:30:30.776
most of these entities [br]will have associated an image
0:30:30.776,0:30:32.490
but not all of them.
0:30:32.490,0:30:37.861
So here there is some missing metadata[br]that could be added.
0:30:37.861,0:30:40.376
You could go to this entity at an image
0:30:40.376,0:30:45.462
and then see first [br]that there is an image missing and so on.
0:30:46.457,0:30:52.047
Another thing is that you could find [br]schema issues.
0:30:52.047,0:30:55.424
For example, if you ask [br]"books by Andrea Camilleri,"
0:30:55.428,0:30:57.711
which is a famous Italian writer,
0:30:57.711,0:30:59.981
you would currently get [br]these three books.
0:30:59.981,0:31:02.681
But he wrote many more.[br]He wrote more than 50.
0:31:02.681,0:31:05.701
And so the question is, [br]are they not in Wikidata
0:31:05.701,0:31:09.704
or is maybe my knowledge[br]not correctly currently like it is.
0:31:09.704,0:31:12.804
And in this case, I know [br]there is another book from him,
0:31:12.804,0:31:14.737
which is "Un mese con Montalbano."
0:31:14.737,0:31:18.207
It has only an Italian label[br]so you can only search it in Italian.
0:31:18.207,0:31:22.103
And if you go to this entity, [br]you will say that he has written it.
0:31:22.103,0:31:27.504
It's a short story by Andrea Camilleri[br]and it's an instance of literary work,
0:31:27.504,0:31:29.220
but it's not instance of book
0:31:29.220,0:31:31.338
so that's the reason why [br]it doesn't appear.
0:31:31.338,0:31:35.904
This is a way to track [br]where things are missing
0:31:35.904,0:31:37.499
in the Wikidata model
0:31:37.499,0:31:39.539
not as you would expect.
0:31:40.794,0:31:42.968
Another reason is just to have fun.
0:31:43.588,0:31:47.546
I imagine that many of you added [br]many Wikidata entities
0:31:47.546,0:31:50.776
so just search for the ones[br]that you care most
0:31:50.776,0:31:52.529
or you have edited yourself.
0:31:52.529,0:31:56.893
So in this case, who developed [br]QAnswer, and that's it.
0:31:56.893,0:32:00.226
For any other questions, [br]go to www.QAnswer.eu/qa
0:32:00.226,0:32:03.575
and hopefully we'll find [br]an answer for you.
0:32:03.782,0:32:05.649
(audience clapping)
0:32:13.994,0:32:17.040
- Sorry.[br]- I'm just the dumbest person here.
0:32:17.530,0:32:22.722
(audience 11) So I want to know [br]how is this kind of agnostic
0:32:22.752,0:32:25.104
to Wikibase instance,
0:32:25.104,0:32:29.020
or has it been tied to the exact [br]like property numbers
0:32:29.020,0:32:31.054
and things in Wikidata?
0:32:31.054,0:32:33.442
Has it learned in some way [br]or how was it set up?
0:32:33.442,0:32:36.456
There is training data[br]and we rely on training data
0:32:36.456,0:32:40.585
and this is also most of the cases[br]why you will not get good resutls.
0:32:40.585,0:32:44.881
But we're training the system[br]by the simple yes and no answer.
0:32:44.881,0:32:48.936
When you ask a question, [br]and we ask always for feedback, yes or no,
0:32:48.936,0:32:51.899
and this feedback is used by [br]the machine learning algorithm.
0:32:51.899,0:32:54.124
This is where machine learning [br]comes into play.
0:32:54.124,0:32:58.600
But basically, we put up separate [br]Wikibase instances
0:32:58.600,0:33:00.482
and we can plug this in.
0:33:00.482,0:33:04.249
In fact, the system is agnostic[br]in the sense that it only wants RDF.
0:33:04.249,0:33:06.618
And RDF, you have in each Wikibase,
0:33:06.618,0:33:08.059
there are some few configurations
0:33:08.059,0:33:10.432
but you can have this on top [br]of any Wikibase.
0:33:11.654,0:33:13.039
(audience 11) Awesome.
0:33:23.573,0:33:27.004
(audience 12) You mentioned that[br]it's being trained by yes/no answers.
0:33:27.073,0:33:32.662
So I guess this is assuming that [br]the Wikidata instance is free of errors
0:33:32.722,0:33:34.356
or is it also...?
0:33:34.356,0:33:37.140
You assume that the Wikidata instances...
0:33:37.140,0:33:40.731
(audience 12) I guess I'm asking, like, [br]are you distinguishing
0:33:40.731,0:33:46.289
between source level errors[br]or misunderstanding the question
0:33:46.289,0:33:50.856
versus a bad mapping, etc.?
0:33:51.706,0:33:55.474
Generally, we assume that the data[br]in Wikidata is true.
0:33:55.474,0:33:59.172
So if you click "no" [br]and the data in Wikidata would be false,
0:33:59.172,0:34:03.023
then yeah... we would not catch[br]this difference.
0:34:03.023,0:34:05.081
But sincerely, Wikidata quality[br]is very good,
0:34:05.081,0:34:08.231
so I rarely have had this problem.
0:34:16.592,0:34:22.068
(audience 12) Is this data available [br]as a dataset by any chance, sir?
0:34:22.209,0:34:27.218
- What is... direct service?[br]- The... dataset of...
0:34:27.218,0:34:30.803
"is this answer correct[br]versus the query versus the answer?"
0:34:30.872,0:34:33.340
Is that something you're publishing[br]as part of this?
0:34:33.340,0:34:38.040
- The training data that you've...[br]- We published the training data.
0:34:38.040,0:34:43.423
We published some old training data[br]but no, just a--
0:34:44.573,0:34:47.313
There is a question there.[br]I don't know if we have still time.
0:34:51.215,0:34:55.104
(audience 13) Maybe I just missed this[br]but is it running on a live,
0:34:55.104,0:34:57.080
like the Live Query Service,
0:34:57.080,0:34:59.393
or is it running on [br]some static dump you loaded
0:34:59.393,0:35:01.690
or where is the data source[br]for Wikidata?
0:35:01.784,0:35:07.014
Yes. The problem is [br]to apply this technology,
0:35:07.014,0:35:08.414
you need a local dump.
0:35:08.414,0:35:10.673
Because we do not rely only [br]on the SPARQL end point,
0:35:10.673,0:35:12.873
we rely on special indexes.
0:35:12.873,0:35:16.192
So, we are currently loading [br]the Wikidata dump.
0:35:16.192,0:35:18.699
We are updating this every two weeks.
0:35:18.699,0:35:20.756
We would like to do it more often,
0:35:20.756,0:35:23.823
in fact we would like to get the difs[br]for each day, for example,
0:35:23.823,0:35:25.271
to put them in our index.
0:35:25.271,0:35:28.719
But unfortunately, right now,[br]the Wikidata dumps are released
0:35:28.719,0:35:31.753
only once every week.
0:35:31.753,0:35:35.150
So, we cannot be faster than that[br]and we also need some time
0:35:35.150,0:35:39.073
to re-index the data, [br]so it takes one or two days.
0:35:39.073,0:35:41.833
So we are always behind. Yeah.
0:35:48.202,0:35:49.780
(moderator) Any more?
0:35:50.430,0:35:53.268
- Okay, thank you very much.[br]- Thank you all very much.
0:35:53.547,0:35:54.966
(audience clapping)
0:35:57.266,0:36:00.165
(moderator) And now last, we have[br]Eugene Alvin Villar,
0:36:00.165,0:36:02.049
talking about Panandâ.
0:36:10.630,0:36:12.637
Good afternoon, [br]my name is Eugene Alvin Villar
0:36:12.637,0:36:15.297
and I'm from the Philippines,[br]and I'll be talking about Panandâ:
0:36:15.297,0:36:18.185
a mobile app powered by Wikidata.
0:36:18.862,0:36:21.678
This is a follow-up to my lightning talk[br]that I presented two years ago
0:36:21.678,0:36:25.004
at WikidataCon 2017[br]together with Carlo Moskito.
0:36:25.004,0:36:26.557
You can download the slides
0:36:26.557,0:36:28.727
and there's a link [br]to that presentation there.
0:36:28.727,0:36:30.868
I'll give you a bit of a background.
0:36:30.868,0:36:33.471
Wiki Society of the Philippines,[br]formerly, Wikimedia Philippines,
0:36:33.471,0:36:37.477
had a series of projects related [br]to Philippine heritage and history.
0:36:37.477,0:36:41.705
So we have the usual photo contests,[br]Wikipedia Takes Manila,
0:36:41.705,0:36:43.238
Wiki Loves Monuments,
0:36:43.238,0:36:46.657
and then our media project[br]was Cultural Heritage Mapping Project
0:36:46.657,0:36:49.094
back in 2014-2015.
0:36:50.044,0:36:53.039
In that project, we trained volunteers[br]to edit articles
0:36:53.039,0:36:54.389
related to cultural heritage.
0:36:54.914,0:36:59.032
This is our biggest, [br]and most successful project that we had.
0:36:59.032,0:37:03.037
794 articles were created or improved, [br]including 37 "Did You Knows"
0:37:03.037,0:37:05.238
and 4 "Good Articles,"
0:37:05.308,0:37:08.688
and more than 5,000 images were uploaded[br]to Commons.
0:37:08.688,0:37:11.039
As a result of that, we then launched
0:37:11.039,0:37:13.689
the Encyclopedia [br]of Philippine Heritage program
0:37:13.689,0:37:18.444
in order to expand the scope[br]and also include Wikidata in the scope.
0:37:18.444,0:37:21.695
Here's the Core Team: myself,[br]Carlo and Roel.
0:37:21.695,0:37:26.870
Our first pilot project was to document[br]the country's historical markers
0:37:26.870,0:37:29.153
in Wikidata and Commons,
0:37:29.153,0:37:34.053
starting with those created by[br]our historical national agency, NHCP.
0:37:34.053,0:37:38.904
For example, they installed a marker[br]for our national hero, here in Berlin,
0:37:38.904,0:37:41.421
so there's no Wikidata page[br]for that marker
0:37:41.421,0:37:45.102
and a collection of photos of that marker[br]in Commons.
0:37:46.166,0:37:50.397
Unfortunately, the government agency[br]does not keep a good database
0:37:50.397,0:37:53.480
up-to-date or complete of their markers,
0:37:53.480,0:37:58.004
so we have to painstakingly input these[br]to Wikidata manually.
0:37:58.004,0:38:02.772
After careful research and confirmation,[br]here's a graph of the number of markers
0:38:02.772,0:38:07.466
that we've added to Wikidata over time,[br]over the past three years.
0:38:07.466,0:38:11.230
And we've developed [br]this Historical Markers Map web app
0:38:11.230,0:38:15.289
that lets users view [br]these markers on a map,
0:38:15.289,0:38:21.051
so we can browse it as a list,[br]view a good visualization of the markers
0:38:21.051,0:38:23.253
with information and inscriptions.
0:38:23.253,0:38:28.885
All of this is powered by Live Query[br]from Wikidata Query Service.
0:38:29.732,0:38:32.005
There's the link [br]if you want to play around with it.
0:38:33.349,0:38:37.428
And so we developed [br]a mobile app for this one.
0:38:37.428,0:38:42.117
To better publicize our project,[br]I developed the Panandâ
0:38:42.117,0:38:45.434
which is Tagalog for "marker",[br]as an android app,
0:38:45.434,0:38:48.393
that was published back in 2018,
0:38:48.393,0:38:53.934
and I'll publish the IOS version[br]sometime in the future, hopefully.
0:38:54.868,0:38:57.892
I'd like to demo the app[br]but we have no time,
0:38:57.892,0:39:00.935
so here are some [br]of the features of the app.
0:39:00.935,0:39:04.586
There's a Map and a List view,[br]with text search,
0:39:04.586,0:39:07.452
so you can drill down as needed.
0:39:07.452,0:39:10.169
You can filter by region or by distance,
0:39:10.169,0:39:12.193
and whether you have marked [br]these markers,
0:39:12.193,0:39:15.499
as either you have visited them [br]or you'd like to bookmark them
0:39:15.499,0:39:16.949
for future visits.
0:39:16.949,0:39:19.482
Then you can use your GPS[br]on your mobile phone
0:39:19.482,0:39:21.860
to use for distance filtering.
0:39:21.860,0:39:26.765
For example, if I want markers[br]that are near me, you can do that.
0:39:26.765,0:39:30.918
And when you click on the Details page,[br]you can see the same thing,
0:39:30.918,0:39:35.850
photos from Commons, [br]inscription about the marker,
0:39:35.850,0:39:40.484
how to find the marker,[br]its location and address, etc.
0:39:41.601,0:39:45.993
And one thing that's unique for this app[br]is you can, again, visit
0:39:46.011,0:39:50.407
or put a bookmark of these,[br]so on the map or on the list,
0:39:50.407,0:39:51.692
or on the Details page,
0:39:51.692,0:39:54.891
you can just tap on those buttons [br]and say that you've visited them,
0:39:54.891,0:39:58.520
or you'd like to bookmark them [br]for future visits.
0:39:58.520,0:40:03.527
And my app has been covered by the press[br]and given recognition,
0:40:03.527,0:40:06.743
so plenty of local press articles.
0:40:06.743,0:40:11.281
Recently, it was selected [br]as one of the Top 5 finalists
0:40:11.281,0:40:15.247
for the Android Masters competition[br]in the App for Social Good category.
0:40:15.247,0:40:17.351
The final event will be next month.
0:40:17.351,0:40:18.999
Hopefully, we'll win.
0:40:20.380,0:40:22.378
Okay, so some behind the scenes.
0:40:22.378,0:40:25.477
How did I develop this app?
0:40:25.477,0:40:28.578
Panandâ is actually a hybrid app,[br]it's not native.
0:40:28.578,0:40:30.745
Basically it's just a web app[br]packaged as a mobile app
0:40:30.745,0:40:32.518
using Apache Cordova.
0:40:32.518,0:40:34.026
That reduces development time
0:40:34.026,0:40:36.181
because I don't have to learn [br]a different language.
0:40:36.181,0:40:37.769
I know JavaScript, HTML.
0:40:37.879,0:40:42.131
It's cross-platform, allows code reuse[br]from the Historical Markers Map.
0:40:42.385,0:40:46.311
And the app is also FIN Open Source.[br]under the MIT license.
0:40:46.311,0:40:49.429
So there's the GitHub repository [br]over there.
0:40:50.469,0:40:53.624
The challenge is [br]the apps data is not live.
0:40:54.750,0:40:56.820
Because if you query the data live,
0:40:56.843,0:41:00.638
it means you pulling around half [br]a megabyte of compressed JSON every time
0:41:00.638,0:41:03.594
which is not friendly [br]for those on mobile data,
0:41:03.594,0:41:06.723
incurs too much delay when starting[br]the app,
0:41:06.723,0:41:13.097
and if there are any errors in Wikidata,[br]that may result in poor user experience.
0:41:14.253,0:41:18.046
So instead, what I did was [br]the app is updated every few months
0:41:18.046,0:41:20.468
with fresh data, compiled using [br]a Perl script
0:41:20.468,0:41:23.037
that queries Wikidata Query Service,
0:41:23.037,0:41:25.678
and this script also does [br]some data validation
0:41:25.678,0:41:30.944
to highlight consistency or schema errors,[br]so that allows fixes before updates
0:41:30.944,0:41:34.735
in order to provide a good experience[br]for the mobile user.
0:41:35.174,0:41:39.274
And here's the... if you're tech-oriented,[br]here's the more or less,
0:41:39.274,0:41:41.644
the technologies that I'm using.
0:41:41.644,0:41:43.976
So a bunch of JavaScript libraries.
0:41:43.976,0:41:46.287
Here's the first script [br]that queries Wikidata,
0:41:46.287,0:41:48.598
some Cordova plug-ins,
0:41:48.598,0:41:53.035
and building it using Cordova[br]and then publishing this app.
0:41:53.763,0:41:55.586
And that's it.
0:41:55.748,0:41:58.164
(audience clapping)
0:42:01.800,0:42:04.072
(moderator) I hope you win. [br]Alright, questions.
0:42:16.286,0:42:17.990
(audience 14) Sorry if I missed this.
0:42:17.990,0:42:21.317
Are you opening your code [br]so the people can adapt your app
0:42:21.317,0:42:24.501
and do it for other cities?
0:42:24.501,0:42:28.516
Yes, as I've mentioned, [br]the app is free and open source,
0:42:28.516,0:42:31.095
- (audience 14) But where is it?[br]- There's the GitHub repository.
0:42:31.095,0:42:33.610
You can download the slides,[br]and there's a link
0:42:33.610,0:42:36.841
in one of the previous slides[br]to the repository.
0:42:36.841,0:42:38.732
(audience 14) Okay. Can you put it?
0:42:42.392,0:42:43.747
Yeah, at the bottom.
0:42:46.577,0:42:49.222
(audience 15) Hi. Sorry, maybe [br]I also missed this,
0:42:49.222,0:42:51.628
but how do you check for a schema errors?
0:42:53.055,0:42:56.007
Basically, we have a Wikiproject[br]on Wikidata,
0:42:56.106,0:43:02.425
so we try to put the other guidelines[br]on how to model these markers correctly.
0:43:02.425,0:43:05.190
Although it's not updated right now.
0:43:06.197,0:43:09.023
As far as I know, we're the only country
0:43:09.023,0:43:12.874
that's currently modeling these[br]in Wikidata.
0:43:13.930,0:43:20.152
There's also an effort [br]to add [inaudible]
0:43:20.161,0:43:22.411
in Wikidata,
0:43:22.474,0:43:25.705
but I think that's [br]a different thing altogether.
0:43:34.056,0:43:35.895
(audience 16) So I guess this may be part
0:43:35.895,0:43:37.725
of this Wikiproject you just described,
0:43:37.725,0:43:42.800
but for the consistency checks,[br]have you considered moving those
0:43:42.800,0:43:46.743
into like complex schema constraints[br]that then can be flagged
0:43:46.743,0:43:50.583
on the Wikidata side for[br]what there is to fix on there?
0:43:52.930,0:43:55.547
I'm actually interested in seeing [br]if I can do, for example,
0:43:55.598,0:44:00.296
shape expressions, so that, yeah,[br]we can do those things.
0:44:04.256,0:44:06.776
(moderator) At this point, [br]we have quite a few minutes left.
0:44:06.776,0:44:09.026
The speakers did very well,[br]so if Erica is okay with it,
0:44:09.026,0:44:11.238
I'm also going to allow [br]some time for questions,
0:44:11.238,0:44:13.407
still about this presentation,[br]but also about Mbabel,
0:44:13.407,0:44:15.498
if anyone wants to jump in[br]with something there,
0:44:15.498,0:44:17.318
either presentation is fair game.
0:44:22.790,0:44:25.639
Unless like me, you're all so dazzled[br]that you just want to go to snacks
0:44:25.639,0:44:27.955
and think about it.[br](audience giggles)
0:44:29.308,0:44:31.179
- (moderator) You know...[br]- Yeah.
0:44:31.953,0:44:34.491
(audience 17) I will always have [br]questions about everything.
0:44:34.491,0:44:37.642
So, I came in late for the Mbabel tool.
0:44:37.642,0:44:40.350
But I was looking through [br]and I saw there's a number of templates,
0:44:40.350,0:44:43.232
and I was wondering [br]if there's a place to contribute
0:44:43.232,0:44:45.564
to adding more templates [br]for different types
0:44:45.564,0:44:47.620
or different languages and the like?
0:44:50.497,0:44:53.683
(Erica) So for now, we're developing[br]those narrative templates
0:44:53.683,0:44:55.566
on Portuguese Wikipedia.
0:44:55.566,0:44:57.856
I can show you if you like.
0:44:57.856,0:45:02.051
We're inserting those templates[br]on English Wikipedia too.
0:45:02.051,0:45:07.017
It's not complicated to do[br]but we have to expand for other languages.
0:45:07.017,0:45:08.236
- French?[br]- French.
0:45:08.236,0:45:10.465
- Yes.[br]- French and German already have.
0:45:10.465,0:45:11.465
(laughing)
0:45:12.002,0:45:13.018
Yeah.
0:45:15.755,0:45:18.287
(inaudible chatter)
0:45:21.756,0:45:24.446
(audience 18) I also have a question [br]about Mbabel,
0:45:24.446,0:45:27.676
which is, is this really just templates?
0:45:27.676,0:45:33.893
Is this based on the LUA scripting?[br]Is that all? Wow. Okay.
0:45:33.956,0:45:37.404
Yeah, so it's very deployable. Okay. Cool.
0:45:38.102,0:45:40.199
(moderator) Just to catch that [br]for the live stream,
0:45:40.199,0:45:42.745
the answer was an emphatic nod [br]of the head, and a yes.
0:45:42.915,0:45:44.648
(audience laughing)
0:45:44.754,0:45:47.203
- (Erica) Super simple.[br]- (moderator) Super simple.
0:45:47.745,0:45:49.819
(audience 19) Yeah. [br]I would also like to ask.
0:45:49.819,0:45:53.386
Sorry I haven't delved [br]into Mbabel earlier.
0:45:53.386,0:45:57.018
I'm wondering, you're working also[br]with the links, the red links.
0:45:57.018,0:46:00.052
Are you adding some code there?
0:46:03.987,0:46:07.970
- (Erica) For the lists?[br]- Wherever the link comes from...
0:46:07.970,0:46:11.595
(audience 19) The architecture. [br]Maybe I will have to look into it.
0:46:11.595,0:46:13.355
(Erica) I'll show you later.
0:46:20.506,0:46:23.221
(moderator) Alright. You're all ready[br]for snack break, I can tell.
0:46:23.221,0:46:24.456
So let's wrap it up.
0:46:24.456,0:46:26.429
But our kind speakers, [br]I'm sure will stick around
0:46:26.429,0:46:27.958
if you have questions for them.
0:46:27.958,0:46:31.179
Please join me in giving... first of all[br]we didn't give a round of applause yet.
0:46:31.179,0:46:33.221
I can tell you're interested in doing so.
0:46:33.221,0:46:34.886
(audience clapping)