Proposal talk:A central repository of all language independent data

Proposals for new projects

For those in support, proposals for new projects are handled on meta, see m:proposals for new projects. Dedalus 21:20, 15 August 2009 (UTC)

I would point out that it's entirely appropriate to discuss here and form the proposal that will be eventually submitted elsewhere - or to agreee to reconvene there. Feel free to use this space for whatever planning you need. :) -- Philippe 21:31, 15 August 2009 (UTC)

I think it would be powerful to get in touch with official organizations responsible to provide statistical data and find a way to import these data automatically in a database linked with WP. Snipre 12:04, 21 August 2009 (UTC)

This idea - which I think of as "wikidata" - could become a driving force for where Wikipedia needs to go in coming years. When I imagine a world in which every single human being can freely share in the sum of all knowledge, I think of a world where the right data is available, at least as importantly as one where I can find text articles, quotes, news etc.

Creating a wiki to illuminate the world of numbers has some huge advantages - data can't be copyrighted and doesn't need to be stored separately to cover multiple languages. Similar presentation tools - tables, graphs etc - can be automated to call on very different number sets. The whole system would be a vast improvement on having graphs which are really just pictures, and which differ from article to article. An example would be oil price data - there are at least 39 graphs of oil prices at wikimedia commons, many of them very out-of-date but still used in articles.

Wikidata would be as fluid as, say, Wikipedia. Some things like census data would be locked, because there is one right answer, but most datasets - for example, global temperature - have multiple datasets and fascinating distinctions between them. Because human knowledge goes beyond knowing one right answer, the structure of wikidata would need to allow for forks, discussions and (inevitably) edit wars.--Travelplanner 10:43, 16 August 2009 (UTC)

Yes please. Move all infobox data to Meta!

This proposal is basically about the information which is included in Infoboxes. I would propose this info is moved to Meta or WikiData and the transcluded back into the various Wikipedias.

A wikipedia in an obscure language would then localise an Infobox headings and immediately every one of that type of infobox would be available on a search in that language.

It would be important to make the info edittable from within the various local WPs.

Infoboxes are useful structured data so this offline data store could also be useful for all kinds of data mining and intelligent searches with outputs in tables rather than single factoids. A number of external websites are starting to mine infoboxes for just this sort of information. It would be nice if Wikipedia itself had the tools to do this.

Because of the danger of errors in this info being widely used I believe there should be a requirement that every single fact in this wikidata site should have a reference.

Perhaps we could start with the Wikispecies data?Filceolaire 15:00, 24 August 2009 (UTC)

Radically different from Wikispecies

As I understand the proposal it is radically different from Wikispecies. Wikispecies is about creating a catalogue of species (hopefully reflecting a mainstream view) placed in a classification (hopefully reflecting a mainstream view, but until recently this was not the case). But anyway, a highly subjective whole, depending on the judgement of the Wikispecies-editors.

This proposal focuses instead on mirroring data that is being provided by some authority. Quite different. - Brya 09:13, 30 August 2009 (UTC)

Impact?

Some proposals will have massive impact on end-users, including non-editors. Some will have minimal impact. What will be the impact of this proposal on our end-users? -- Philippe 00:04, 3 September 2009 (UTC)

Language issue

This proposal has already been made for wiktionaries, and I strongly opposed it. I even gave the argument that, obviously nobody had proposed this idea for Wikipedias (I hope that I'm not at the origin of this proposal).

Currently, data can be shared: bots can import data from another wikipedia, it's easy when templates are used. This is the best solution.

Adopting the proposal would introduce two classes of contributors: those speaking English well, and other ones. There is no solution to this issue. The issue already exists on Commons, but it's much less serious, because there is much less need for discussions on Commons.

I consider that this would be catastrophic for wikipedias. Why does this issue tend to be forgotten, or to be considered as of little importance? Because everybody participating to discussions such as this one on strategy.wikimedia.org can speak English, other ones don't dare to participate. And the conseqquence of adopting this proposal would be the same. Lmaltier 11:41, 23 January 2011 (UTC)