How will the Wikidata [1] project impact Wikipedia?

Question

Denny Vrandecic · Accepted Answer

How will the Wikidata [1] project impact Wikipedia?

In more ways than we currently imagine, I hope. The first and obvious change is that many infoboxes will become more up to date, especially in the smaller Wikipedias. At the same time, Wikidata should reduce the maintenance cost for the articles having these infoboxes, because the data comes from a central source.

Some non-obvious ways: due to the reduced bot editing the histories of the articles in smaller Wikipedias will not be as flooded by bots as they currently are. Try it out. Go to a small Wikipedia. Take a random article. Look at the edit history. How many bots do you see? How many humans? Making the bots become less common and humans more pronounced could maybe increase the sense of community and thus increase the edit activity in those Wikipedias.

Another non-obvious way: smaller Wikipedias today can hardly afford to create articles for all cities, species, etc., because there are not enough editors to maintain these articles. Wikidata could allow them to create stubs for these articles including informative infoboxes. This in turn might drive more readers to these Wikipedias, which in turn might increase the number of editors.

Also, using Wikidata should decrease Wikipedia's reliance on complex templates in the articles themselves (especially by hiding the huge template call at the head of many pages). This will help with enabling the upcoming Visual Editor to deal with a larger portion of the page quicker. Editing the data in Wikidata should be easier than editing it in a Wikipedia template call. This in turn increases usability of Wikipedia for editors, especially casual ones, and thus might increase the editorship as well.

In short, I hope and expect that Wikidata will increase the availability of knowledge through Wikipedia, increase the number of editors globally, increase the reach of Wikipedia, and increase the quality of Wikipedia. But who knows what will really happen.

Will it change the way people add data to Wikipedia?

If you mean "Will they stop adding data through the template syntax and instead start adding data through a form-based interface", my answer is, I sure hope so. If you mean "Will it change the way people add information to Wikipedia?" I do not expect a big change, actually. Most of the information is and will remain in the text anyway, and this won't be influenced by Wikidata.

Will it enforce structure?

No, in no way will it enforce anything. It is an option that any Wikipedia community can decide for any single page, whether to use some or any of the information offered by Wikidata.

How will it impact projects that depend upon Wikipedia, like DBpedia?

It should improve greatly the extraction of structured data from Wikipedia. DBpedia is an awesome project consisting of several parts, like the extraction of data from infoboxes and other sources, a SPARQL endpoint to the extracted data, several different visual interfaces to the extracted data, etc.

Wikidata will strictly simplify the extraction process, thus freeing up resources currently spent on the extraction. All parts of DBpedia will improve by using the higher quality dataset that Wikidata will hopefully provide. In short, it should strongly and positively impact these dependent projects.

Nizil Shah · Answer

In simple explanation:
1. Removal of complex wiki markups and templates which make new editors confused.
2. Up to date data as central repository for structured data: eg. When new data on population available for city, change in central repository will reflected across all language Wikipedias. Less maintained by bots and editors %3E huge saving of time and source
3. Automatically updated lists:
When new president elected there will be no need to edit lists of presidents on all Wikipedia lists, it will be automatically updated based on data on wikidata. List on cities by population will be updated automatically by new data.
4. Better infographics would be created from data like railway lines, family trees, hierarchy etc based on wikidata
5. Multiple structured data can be available on wikidata for same thing based on qualifiers and sources which will be available on Wikipedia unlike current infoboxes without sources
6. Large number of Stubs on smaller wikipedias derived from data from wikidata
7. Better linked Wikipedia to other Authority controls
8. All Wikipedia sister projects connected better
9. Less available data like Geo coordinates will be available across all Wikipedias
10. Large range of data related to item available unlike lnfoboxes which have certain properties only
11. In short change of small data at one place reflected across large spectrum

Wikidata is still in infancy, it would change the way data accessed on Wikipedia and across the internet or sematic web..

It will serve same as what wikimedia commons serve for media..

Daniel Kinzler · Answer

There are four main differences:

* Wikidata provides data to Wikipedia, while DBpedia extracts data from Wikipedia.
 * Wikidata's ontology is community curated and part of the data maintained on the site, while DBpedia’s ontology is statically defined, and much more strict.
 * Formally, Wikidata only asserts statements (who claims what), while DBpedia asserts facts, often causing contradictions.
 * Wikidata is licensed CC-0, and is this re-usable without any restrictions, while DBpedia is licensed CC-BY-SA, which requires author attribution - which is a good thing generally, but impractical for a knowledge base automatically derived from text.

Mark Hetherington · Answer

Unlike DBpedia, Wikidata is a project of the Wikimedia Foundation.

It is directly integrated with Wikipedia, storing in particular all the inter-language links between corresponding articles on different language editions of Wikipedia. Wikipedia templates can automatically pull data values from Wikidata, too, with easy access to the Wikidata entry corresponding to the current article.

DBpedia is a third-party project with no integration on Wikipedia's side.

Mark Hetherington · Answer

I'll answer this one indirectly. This is Georges Seurat's Un dimanche après-midi à l'Île de la Grande Jatte:

It's a very well-known painting and if you've ever heard of pointillism [ https://en.wikipedia.org/wiki/Pointillism ], you've probably seen it somewhere or other (I hope to see it in person next time I visit Chicago). Pointillism takes many, many tiny dots of colour, and in aggregate, they are resolved into a coherent scene.

Facts are like dots of colour. We can specify them with a high degree of precision. Imagine a digital colour-dot, and we can resolve it to data: location (X and Y coordinates), colour (take your pick of HSL, RGB, CMYK, etc.), transparency, radius … and so on.

Nonetheless, we can't reduce the whole painting that way, not without obfuscating the core thing, the scene that the artist wanted us to see. Imagine a small book filled with XML describing each dot in sequence. There's something lost there; the average person couldn't hope to comprehend it without assistance or a long analysis.

Wikidata is for dots: facts that we can nail down and measure. These dots are really useful, because it's not terribly hard to connect a few of the most important ones and get an image.

Wikipedia is for the painting. Wikipedia is the place where we'd gloss over the individual dots and give an impression of the trees, the water, the people relaxing on the grass, and the slowly lengthening shadows of the afternoon. We can't break those down to dots… that'd be like trying to define concrete edges for the diffuse tree foliage in the upper-left of the painting. We can highlight it in context: can put collections of dots together and call them images of a man sitting and holding his cane, or a woman with an umbrella, or whatever the big picture may happen to be.

Andreas Kolbe · Answer

As others have said, no. There are data common to both – naturally so, as a lot of the data in Wikidata has been imported from Wikipedia – and data only found in Wikidata or Wikipedia.

It may interest you that Wikidata is quite controversial within Wikipedia. See for example the ongoing discussion here: Wikipedia talk:Wikidata/2017 State of affairs - Wikipedia [ https://en.wikipedia.org/w/index.php?title=Wikipedia_talk:Wikidata/2017_State_of_affairs&oldid=800636865 ].

Wikipedians’ concerns are focused on –

* the fact that Wikidata lacks policies for reliable sourcing – unlike Wikipedia, Wikidata does not require users to cite a reliable source for content added;
 * the lack of Wikidata policies ensuring that living people are not defamed: to give an example, from May 2013 to June 2015, Wikidata said about Larry Silverstein that "In May of 2013, Larry and his pal Tom, confessed to the implosion of the WTC, and later were sentenced to multiple life terms in prison" (this was of course a complete lie) instead of describing him as an "American businessman" – you can see the relevant Wikidata edits and the time when they were made here: Difference between revisions of "Larry Silverstein" [ https://www.wikidata.org/w/index.php?title=Q463668&diff=43216372&oldid=27579105 ], Difference between revisions of "Larry Silverstein" [ https://www.wikidata.org/w/index.php?title=Q463668&diff=221802840&oldid=172224477 ];
 * Wikidata’s apparent inability to spot and remove vandalism in a timely manner. For example, Wikipedians are concerned about vandalised Wikidata descriptions being prominently displayed to mobile users reading or searching for a Wikipedia article.
While the Wikimedia Foundation is making efforts to integrate Wikidata content into Wikipedia, especially for mobile viewers, this is strongly resisted by many established Wikipedians – in part because a large part of Wikidata content has been automatically imported from Wikipedia (including lesser language versions of Wikipedia that have less stringent quality controls than the English Wikipedia) and because mass imports are being done without regard for whether the imported data is reliably sourced in the relevant Wikipedia article or not.

For more on this topic, see this December 2015 op-ed in Wikipedia’s community newspaper: Whither Wikidata? [ https://en.wikipedia.org/wiki/Wikipedia:Wikipedia_Signpost/2015-12-02/Op-ed ]

Wikipedia has its own well-known reliability problems, but Wikidata’s quality control systems seem to be far less efficient even than Wikipedia’s.

Pete Forsyth · Answer

For the English Wikipedia (and, I’m almost certain, for the German Wikipedia), the answer is that most but not literally all, Wikipedia articles have a corresponding Wikidata item. I don’t know how true it is for other language editions.

When Wikidata launched a few years ago, automated processes (scripts or bots) “scraped” information from Wikipedia to populate the new database. It may be that at that point in time, every Wikipedia article had a corresponding Wikidata entry added; or, at least, most did.

But when you compose a new Wikipedia article, it does not get automatically added to Wikidata. There may be some scripts that crawl around looking for new articles, but I doubt they get every single article.

Since it’s closely related to your question, let me point out two areas where I find Wikidata useful these days:

(1) It’s useful to provide a solid link among entries on various Wikimedia projects - e.g., Wikipedia, Wikimedia Commons (image and multimedia repository), Wikisource (republication of public domain texts, like old books), and Wikiquote (noteworthy quotations). Here’s an example: [George W. Joseph [ https://www.wikidata.org/wiki/Q386710 ]] At the bottom, note that there are links to Wikipedia, Wikisource, and Wikiquote; and there is also an image from Commons included. The existence of this Wikidata page serves, among other things, to populate the “authority control” box at the bottom of some of these pages, like here: [G. W. Joseph’s “author page” on Wikisource [ https://en.wikisource.org/wiki/Author:George_W._Joseph ]] which gives you easy access to entries about the guy in all kinds of online databases.

(2) If a person or topic is not quite notable enough for a Wikipedia article (or if proving notability will require a bunch of research that’ll take a while to get around to), Wikidata can be a convenient place for a Wikipedia writer to “park” some information so you can easily get back to it in the future. In the meantime, it makes that information a bit easier for others to find on the web. Here’s an example: Charles Clarence Chapman [ https://www.wikidata.org/wiki/Q52161570 ] I found a couple sources that give info like his full name, the job he was best known for, and his birth date; now you can find some of that stuff too, a bit more easily than I did.

For a more authoritative answer to your basic question, I’d suggest reaching out to Lydia Pintscher [ https://www.quora.com/profile/Lydia-Pintscher ] or User-13692603739530639813 [ https://www.quora.com/profile/User-13692603739530639813 ].

Daniel Kinzler · Answer

Wikibase can be installed and used as a MediaWiki extension without too much trouble, see Wikibase/Installation - MediaWiki [ https://www.mediawiki.org/wiki/Wikibase/Installation ]. We took care to not tie Wikibase to the Wikimedia environment, though admittedly, your focus is to power Wikidata, and features that would only benefit 3rd party users have low priority.

Some limitations you should be aware of:

While the “repository” functionality should work fine on a 3rd party wiki, the “client” functionality only works for local wikis. So, you cannot install the client part of Wikibase to access information from Wikidata, like Wikipedia does. You can only make your client wikis use information from your own Wikibase repository, which runs in your local network. We plan to change this some day, but it’s pretty far down on the todo list.

Similarly, it is not yet possible for your repository to re-use the vocabulary (properties and items) defined on Wikidata. We are currently working on making this kind of “federation” possible within the local network. Support for “remote“ federation will take a while.

If you choose to set up a wiki family with a Wikibase repository and several client wikis, you will need to configure “change propagation”. This essentially means that you need to run the dispatchChanges script regularly, typically from cron.

Depending in your database setup, you may run into some complications, though: Wikibase only runs on MySQL (or MariaDB), and if you are using unicode/utf-8 encoding in your MediaWiki database (as opposed to binary), you may run into errors concerning the index key length. You may have to manually change the table definitions and use truncated keys.

If you decide to put data items into the main namespace, like one Wikidata, make sure there are no other pages in that namespace. In particular, you have to move the wiki’s main page to another namespace first. I recommend to move it to the Project namespace.

Overall, the 3rd party experience could use some polishing, but you should be able to at least set up a Wikibase repo and collect structured data.

Daniel Kinzler · Answer

On any page on Wikipedia that is connected to a data item on Wikidata, you can find a link to the data item in the sidebar on the left, in the "tools" section. It says "Wikidata item".

The information that was moved to Wikidata would usually still be visible on Wikipedia. To edit it, you currently have to go to wikidata. Editing it directly from Wikipedia is on the road map, but will take a while. The "mental model" of an infobox is very different from how we represent knowledge on Wikidata, so designing a godd editing interface is quite tricky. And whatever we do, we have to stay compatible with "old style" infoboxes too, which is tricky already.

Tom Morris · Answer

Some preliminary analyses have been done and some basic tooling created, but not much else.

The idea that all or even a significant portion of Freebase data will migrate to Wikidata is mostly a myth (that Google is happy to let propagate). Google will make some unknown portion of the Freebase data available.  Wikidata is then free to reject as much as they like, up to and including all of it.

Although it was originally thought that Wikidata wouldn't have the same draconian "notability" rules as Wikipedia (where "notable" is defined from the point of view of a white Western 25-year old male computer geek - Darth Vader rules!), it turns out that notability rules are very much alive at Wikidata (in addition to them not wanting large volumes of data for performance reasons).

The former Wikidata project leader (currently employed at Google) is on record as saying:

%3E  "16:51:09 %3Cdennyvrandecic%3E so, in short, Freebase will not add that incredibly much to Wikidata as it is generally assumed"
 #wikimedia-office log [ https://tools.wmflabs.org/meetbot/wikimedia-office/2015/wikimedia-office.2015-03-31-16.00.log.html ]

Of course, I guess another way to look at this is that the "migration" is nearly complete. :-)

Personally, I expect the vast majority of the data in Freebase which is not already part of Wikidata to end up on the cutting room floor.

Ayebilla Avoka · Answer

Try to run this search on your internet browser and see if you can get any good outcome: “names of American politicians whose parents were politicians & journalists,born outside of the US.” If you're lucky so as to get results, it will be one at a time and you need to click each and every link, read through the content to find the names you want (the article may even be one that is not entirely about the person, which means it may be a mere mention of a similar name or just one of your search key words and you waste your time reading). However, this can be done with just a few lines of code(max 5) and you have all the of the names provided they're on wikidata!(easy life, right?).

The problem is, the current database structures do not allow cross-querying of data! Wikidata is the solution! It is a database that allows supports cross-qurrying of data provided the data is stored in it.

If you and I take the pain to teach wikidata about things, we save ourselves and others a great deal of work having to comb through the internet looking for such precise information! Students and researchers will have an easy time doing research as they will save time and energy and data having load different pages just to find little information.

Lydia Pintscher · Answer

MediaWiki is software. It is what runs Wikipedia, Wikimedia Commons and many other projects out there. It can be extended using a wide variety of extensions. One of those extensions is called Wikibase. It gives MediaWiki the ability to store structured data. It is what is used for example on Wikidata.

So to sum it up: MediaWiki is software. Wikidata is a content project like Wikipedia. To run Wikidata we use MediaWiki plus certain extensions.

Pete Forsyth · Answer

Yes. There are probably many kinds of examples, but here's one: Context might make information about a private individual important to a Wikipedia article, but there might be a moral reason to keep that information constrained to the article as much as possible.

For instance, imagine that a popular politician campaigns heavily on the issue of drunk driving. But her husband is discovered late in the campaign cycle to be a repeat drunk driving offender. Her once sure-thing victory is undermined, and she loses office.

Her husband's drunk driving habit would need to be mentioned in her Wikipedia bio, as it impacted an important career. But her husband's name is only publicly known because of his association with her; there's no compelling public interest in keeping *any* structured information about him in a public database -- not even his name, much less his drunk driving habit. (Or if there is, that's a matter for the state and criminal justice system to engage with, with its own accountability structures -- not something to be decided by wiki policy.)

Lydia Pintscher · Answer

Wikidata is coming along great. Our community and content is growing and our content is being used more and more to build cool things inside and outside Wikimedia.

At Wikimania 2017 I gave a talk about the current state of Wikidata and what is coming. The video is here [ https://www.youtube.com/watch?v=P2e9zRWCKuw ] and the slides are here [ https://commons.wikimedia.org/wiki/File:Wikimania_2017_-_The_(Wiki)Data_(R)Evolution.pdf ].

Sorina · Answer

Hello,

Thx for the A2A: Andrew Gorh [ https://www.quora.com/profile/Andrew-Gorh ]

A small hint at:

* http://www.opensym.org/wp-content/uploads/2017/08/a1-piscopo.pdf
 * Is Wikidata up to date with Wikipedia? [ https://opendata.stackexchange.com/questions/9858/is-wikidata-up-to-date-with-wikipedia ]
 * http://korrekt.org/papers/Wikidata-RDF-export-2014.pdf
Hope it helps.

Lydia Pintscher · Answer

You can find a list at the bottom of our dashboard here: Grafana [ https://grafana.wikimedia.org/dashboard/db/wikidata-entity-usage ] So right now the ones who make use of Wikidata the most are English, Swedish, French, Dutch and Chinese.

Nahid Hossain (নাহিদ হোসেন) · Answer

I am focusing on wikidata for about 3 years now. I am learning the new world of data structures.

Wikidata is restructure database system. This kind of table data used in a program, machines or softwares.

The biggest use of wikidata is in search engines. As this is the open licensed database, anyone can use this data.

Students works on their pilot test projects using the wikidata databases.

Andre Engels · Answer

Almost all Wikipedia pages have a corresponding Wikidata subject. However, in most cases just a small part of the actual information from Wikipedia is available on Wikidata (and sometimes the opposite is true as well). The Wikidata subject may be no more than the name in English, a small description in case there is more with that name and a link to the Wikipedia page.

Amir E. Aharoni (אמיר א׳ אהרוני) · Answer

MediaWiki is a piece of software. It's a content management system [ https://en.wikipedia.org/wiki/Content_management_system ] that is used for many different wiki sites, among them Wikipedia, Wikidata, and thousands of others. It's installed on a web server. It has its own website, mediawiki.org [ http://mediawiki.org ], which runs on MediaWiki as well.

Wikidata is a website, which uses the MediaWiki software with some extensions to collect and display pieces of data about things around the world. This data can later be used in Wikipedia and other websites.

Ken Lawton · Answer

You should probably ask Jimmy Wales [ https://www.quora.com/profile/Jimmy-Wales ], as he appears to be a member here.

But my understanding is that they use the money to keep the lights on. Running a website the size and scope of Wikipedia, and all the others, means a lot of people, server space, offices, and all the rest.

If we were to go solely on faith that their requests for donations are legitimate, they are barely scraping by in their non-profit capacity. And it is for this reason that at least once a year they do a donation drive.

For my part… I use Wikipedia ALL the time. I mean at least 3–4 times a day, I go and look stuff up. So when they ask for donations, I throw them a few bucks. It’s the least I can do. And considering all the information and entertainment I’ve gathered from the site, I think it’s well worth it.