#rC3 - Introduction to Wikidata

rC3 Wikipaka intro music

Léa: Hi, everyone, I'm Léa. Here's
Mohammed, and we're going to introduce you

to Wikidata today.
Mohammed: Yes, hi everyone. So in the

course of the talk, if you do have a
question, just feel free to ask them in a

chat. And then we are going to try and
answer them at the end of the talk. Yes.

So let's dive straight in. What is
Wikidata? Wikidata is a free knowledge

base that is based on facts and references
that anyone can edit and reuse. It is part

of the Wikimedia projects. And like all of
us, to start open projects, Wikipedia is

multilingual and has no language barriers.
Data in Wikidata is released under CC0

license. That means Wikidata's data is in
the public domain and it has no exclusive

intellectual property rights that is
applied to it. Wikidata is not a primary

source of information. It only aggregates
or collects structured data that is

already available, some of which are links
to other databases. So it is not meant to

be a place for original research. Wikidata
is made for humans and machines, and is

available for everyone to use, whether on
other Wikimedia projects or outside of it.

Next slide. So what is in the Wikidata?
Wikidata was launched some eight years ago

and was originally created to solve the
problem of unstructuredness in the plain

text format that information in Wikipedia
is rendered in, and also to provide a

central storage location where all of the
different language Wikipedias can connect

and talk to each other. Today, Wikidata
has since outgrown its intended purpose

and has become so big and successful that
it is not only, you know, the most edited

Wikimedia projects, Wikidata's data is now
used more outside of the Wikimedia project than

within it. There are more than 25,000
active editors. That means people who make

at least one edit every month. Wikidata
is used across 800+ Wikimedia projects in

more than 300 languages. And it's
interesting to note that the largest

proportion of Wikidata's items are in the
category of scholarly items comprising

about 30% of the whole. Next slide. So
far, people and bots have made more than

1.3 billion edits to Wikidata and created
more than 91 million items. This map you

see here is a visual impression of
geolocated items currently existing on

Wikidata. So, the bright areas are items
that have coordinates, location property

added as a statement. Next slide. So
Wikidata has a vision, and what is this

vision? Wikidata's vision is to give more
people more access to more knowledge. So

Wikidata gives access to information,
regardless of the language that people

speak, because Wikidata is multilingual,
it expects translations of so-called Q

numbers into different languages. And so
doing Wikidata helps us support the

smaller Wikimedia projects better, you
know, by helping them to benefit from all

of the work that the bigger projects are
doing. And applications and projects

outside of Wikimedia are also able to
benefit from the rich datasets in

Wikidata. So in a nutshell, Wikidata can
be thought of as an online repository of

structured data that anyone can edit and
reuse. Next slide. OK, now, how is

Wikidata connected to Wikipedia and other
Wikimedia projects? Among other things,

Wikidata can assist sister projects
with more easily maintainable infoboxes.

So the table at the right corner of this
article on Wikipedia is called an infobox,

which I'm sure you've seen before, and
Wikidata is able to retrieve content on

Wikidata into those infoboxes [distorted].
And for smaller language Wikipedias like,

you know, Catalan Wikipedia or Welsh
Wikipedia that, that readily leverages

Wikidata to see their content. And it is
helpful because it's, it helps to reduce

editing workload for volunteers. Next
slide. So what should we expect to see on

a typical Wikidata item? Wikidata
expresses relationships in the form of

triples that use items starting with "Q"
and property starting with "P", OK, and

the item will typically be made up of at
least one statement. So in this example

you see on the screen we have two
statements about an entity called Douglas

Adams. The first statement, Douglas Adams
was educated at P69 St John's College.

What this means is that this statement is
qualified by further properties. That is

the academic major, academic degree, the
start time and then the end time and

qualifiers add more meaning to statements.
So Wikidata records not just statements,

but also their sources. And as you can see
here, this helps us to reflect the notion

of verifiability on the project so that
statements Douglas Adams was educated at

St. John's College has two open references
that points to the source of that

information. And the second statement,
Douglas Adams, Q42, was educated at P69,

Brentwood School, only has the qualifiers
start time and end time, and it has no

references, so a single statement consists
of a property that is made up of a value

with or without a reference or with or
without qualifiers. Next slide. So a

typical Wikidata item looks like this, and
you can edit by clicking on the edit

button, it has this pen symbol with edit
next to it. As you can see, each item has

a unique ID that is Q followed by some
number. In this case, the item Douglas

Adams has QID of Q42. And when you look at
the top, there's a termbox. We call it, we

call it the termbox at the top, at the
top, that contains the label in different

languages. A description of the items that
is more of a short phrase telling us what

the item represents. It's says here in
English that Douglas Adams is an English

writer and humorist. Then there is the
alias next to the description which, aside

from the label, tells us what the item
could also be known by here. Next slide.

So, creating a new item is as simple as
going to any page on Wikidata and clicking

on create a new item. And once you click
on create a new item, you get to fill in

the form that is asking for a label,
description and an alias and QIDs are

assigned automatically. Next slide. Next
slide. Next slide, please. Alright, so

there are tools that allow us to edit
Wikidata more efficiently and make bulk

edits to Wikidata, such as Quick
Statements and OpenRefine. Please go to

the previous slide. OK, yeah, right, so,
yeah, Quick Statements and OpenRefine

allow us to make automated edits and
changes to Wikidata. Other tools are

available that allow us to visualize
Wikidata's data. Some of them enhances the

user interface of Wikidata, and these
could include scripts that editors can

install or they could be gadgets that may
be enabled in your preferences settings.

Next slide.
Léa: Alright. So, um, so far, Mohammed

told you about how we describe concepts in
Wikidata, and that's what we've been doing

for the first years of the project, but in
2018, we also started storing a new type

of information in Wikidata, which is
lexicographical data, which is basically

information about words and phrases in all
kinds of languages. And so you see on the

left the data model that is a bit complex
and that's why I'm not going to get too

much into details now but we can talk
about this later. And you can see an

example on the right where we basically
describe the word "Luftballon" in German

and we indicate the language, the lexical
category and all kind of informations that

are not about the object any more, but
actually about the word and how it's

composed of two words, as we like to do in
German and things like this. So, again, if

you want to know more about this, you can
have a look at lexicographical data in

Wikidata or we can talk about it together
later in the questions, for example. So

Wikidata doesn't come alone, it comes with
a bunch of tools that have been, some of

them have been developed by the
development team of Wikidata, some of them

have been developed by the community
themselves in order to do things more

efficiently. That can be, for example,
adding data and some of the tools have

already been mentioned by Mohammed, that
can also be matching data with other

databases, querying the data, reusing the
data. There are also a bunch of tools that

are about watching the data and watching
its quality, watching what edits have been

done recently and so on. And you can find
the page that is called Wikidata Tools on

Wikidata to discover plenty of these tools
and you can, of course, create your own.

So we mentioned that the goal of Wikidata
is to be reused by everyone, but you may

wonder who is actually reusing the data.
Well, the first reusers of Wikidata's data

is actually the Wikidata community itself,
the Wikidata editors, because all of these

items are connected. So one item can be
linked from another, the content of one

item can be reused on another and so on.
The Wikimedia project such as Wikipedia,

but not only. Wikimedia Commons,
Wikisource, almost all of the Wikimedia

projects at that point reuse part of the
data that is coming from Wikidata, and

then we have companies, from the biggest
ones to the small ones because the data is

in CC0 everyone can just reuse the content
that they need. We have, of course, public

institutions such as museums, libraries
and so on. We also have journalists and,

for example, data journalists. We have
scientists and researchers and probably

much more. And the thing is that we don't
necessarily know who's reusing the data

because it's here in the open but there
are probably many usages that we don't

even imagine. So if you're using Wikidata,
or if you would like to use Wikidata's

data, let us know, because we are always
interested to discover more. Now, the

question is: How can one reuse Wikidata?
I'm going to present very quickly one of

the most popular way to query the data.
I'm not going to get into details right

now because there will actually be a
workshop at the conference in two days on

day three about the query service so I'm
gonna let you go there and discover more

about how to use it. The query service is
basically a SPARQL endpoint, SPARQL being

a query language where you can basically
ask questions to Wikidata and get lists or

visualizations as a result. For example,
here's the map of the airports of the

world named after the person and the color
of the dot, it represent the gender of the

person. Or you can make a list of country
flags that are including a sun, because if

the data is properly modeled in Wikidata,
you're able to describe, what are the

different elements that compose a country
flag? Or you can have this bubble charts

with the occupation of accused witches,
because why not? That's the kind of data

we have in Wikidata. Now, there are other
ways, of course, to query the data, I'm

not going to get into details right now,
but if you want to talk more about this,

you can, for example, join the Wikidata
meetups that are gonna happen tomorrow. We

have dumps of the data where you can
download part of or all of the data in a

file. We have a bunch of APIs to access
the data directly from your program. And

on a Wikimedia project specifically, the
community developed a bunch of templates

that are using Wikidata's data using Lua.
And now for something a bit different,

Wikibase. You may have heard of it and you
may even have wondered, OK, what's the

difference between Wikibase and Wikidata?
Well, Wikibase is basically the software

powering Wikidata and, more precisely, the
MediaWiki extension that is turning

MediaWiki into a database. And so,
Wikibase was started to power Wikidata

but it also started developing on its own.
Wikidata is still for now the biggest

existing Wikibase instance, but people can
also install Wikibase directly on their

server and basically create their own
little personal or public Wikidata. And

the development is still ongoing, there
are all kind of super exciting features

coming up soon. And, for example, the
ability to connect better Wikidata and

your own instance of Wikibase, for
example, to be able to reuse data that is

already in Wikidata and to connect it to
the data that you have in your own

Wikibase. So, if you're interested in
Wikidata, if you want to know more, there

are a bunch of pages that you can find.
There is a help portal, the Project Chat

is the main discussion page on the wiki
where you can interact with the other

editors, the community. It's super
important to get in touch with them if you

want to get started with Wikidata. We also
have a mailing list. We have a newsletter

that is called Weekly Summary that you can
find on wiki but also if you subscribe to

the mailing list, you will also receive
it. And then we have some accounts in the

social media, on Twitter, there is a
Facebook group, there is a Telegram, um,

that is linked from the Project Chat and
there is also an IRC channel. So you can

basically find people from the Wikidata
community everywhere. So we are

approaching the end of the session, but
it's not done, we have more Wikidata

related sessions at the c3 in the
Wikipaka. So, for example, tomorrow you're

going to get an introduction to Wikidata,
specifically for journalists and

especially data journalists. Then in the
afternoon, we're gonna have two Wikidata

meetups. The first one is gonna be in
German. The second one is gonna be in

English. So depending on your preferred
language, you can attend one or the other

or both, and on day three, as I mentioned
before, we're going to have a workshop to

learn how to query Wikidata's data with
SPARQL. So feel free to have a look and

check them also in the main schedule of
Wikipaka. Thank you very much for

attending this session. These are our
contact details if you want to, to contact

us. And of course, you can now ask
questions, as we mentioned in the chat or

with the hashtag. And we will be very
happy to answer all your questions right

now.
Herald: Thank you for your input and the

overview about Wikidata. There has been a
few question or questions already answered

by Joel in the IRC channel. One was about
the big dump of scholarly data and what

scholarly data is and how this came to be
in Wikidata. But there is one more

question from the chat right now Till asks
can I add new types of data that are not

yet tracked in Wikidata?
Léa: So I'm wondering, what do you mean

exactly by type of data? Maybe you can
give a bit more details because that can

mean a lot of things. Wikidata, the data
model of Wikidata is very flexible and

it's absolutely not set in stone. Every,
every week the community comes up with

some new ways to describe things.
Sometimes we realize that there is an area

of the world that we completely forgot to
cover, and then we create new properties

to describe, for example, a certain type
of, I don't know, of concept, a certain type

of building or objects that we or
philosophical concept that we didn't

describe yet. So this is always in
movement, in action. When it comes to what

we actually call data types, which is, for
example, a string of text or a date or a

picture, we have all kind of data types
like this, this is a bit more complicated

and overall, it's quite rare that we add a
new data type and it needs a strong, like,

use case so we add that to the software. I
hope that it answer your question and if I

didn't, feel free to ask again.
Herald: Yeah, we've got a feedback. The

example Till meant was, there's a, there's
an organization or a project called

Parliamentwatch in Germany. There was one
talk earlier today where they try to track

and scrape and analyze the parliamentary
protocols. And one big issue they had was

with structural data about all the members
of parliament and how they are organized

and stuff like that. And, um, well, if I
remember correctly, there actually was a

project that tried to include the
structural data of of members of

parliament in Wikidata, if I'm not
mistaken.

Léa: Absolutely. It's a WikiProject
that is called, um, something politicians,

all politicians. I don't remember the
exact name right now, but indeed. Some

people are already working on members of
parliaments and, like, political people in

general. So it's very likely that there is
already a way to structure the data. The

best way is to contact the people directly
involved on this, on this WikiProject.

WikiProjects, by the way, are pages where
basically people who have a specific topic

of interest gather and can discuss about
the specific questions about the topic.

Um, so have a look at this, at this
project about politics and, um, yeah. Try

to see if, if anything is missing, but
generally Wikidata definitely welcome

information about about politicians, about
member of parliaments, this kind of stuff.

What we do not do, however, is store the
full, like, documents, for example, in

that case, the reports or the documents,
that belongs elsewhere. Maybe on Wikimedia

Commons, for example, if it's possible, if
the license allows it. But on Wikidata,

we'll be happy to store the metadata about
them.

Herald: Alright, Joel just posted the link
to the WikiProject, Every Politician, so

if anybody looks for Every Politician on
Wikidata, they will find the project. So

basically, the bottom line is pretty much
anything is possible in Wikidata, right?

Léa: Yeah, thank you Joel, and hi. Almost
everything. So on Wikidata, just like on

Wikipedia, we still have some criteria to
define what can get in Wikidata and what

not, because we are aware that this
knowledge base, it needs to stay quite

general and it cannot contain absolutely
everything. For example, the community

decided a while ago that they would not
create one item for each human living or

who used to live on Earth, that's just not
possible, so there are some notability

criteria that you can find in the help
pages and I would say that the level of,

like, how fine-grained the data should be has
to be discussed with the community and the

good thing about having Wikibase also
available as a separate instance of

Wikidata is that if some people want to
work on a topic where they have some

information that is very, very specific
and would maybe not fit the scope of

Wikidata, they can create their own
Wikibase and then they can connect the

content with what is already in Wikidata.
So altogether, in this Wikibase ecosystem,

yes, pretty much everything is possible.
Herald: Well, the future is certainly

here, at least, with Wikidata. Thank you
again, Léa and Mohammed, for your

insightful introduction to Wikidata and
we're looking forward to more people

joining you in your efforts. Thanks for
your presentation.

Léa: Thank you. See you soon.

rC3 Wikipaka outro music

Subtitles created by c3subtitles.de
in the year 2021. Join, and help us!