Proposal:Wiki language tools

Status (see valid statuses)

The status of this proposal is:
Request for Discussion / Sign-Ups

Every proposal should be tied to one of the strategic priorities below.

Edit this page to help identify the priorities related to this proposal!


  1. Achieve continued growth in readership
  2. Focus on quality content
  3. Increase Participation
  4. Stabilize and improve the infrastructure
  5. Encourage Innovation


Summary

That a new project Wiki-language tools along the lines of "Google language tools"

Proposal

That the foundation should create a language tool that can serve to translate a wikiarticle or wiki page in any one language into any other, that the tool be designed in such a way that it can be taught by users to create ever better and accurate translations, and that it be so structured that new languages can easily be added.

Motivation

The genesis of this proposal lies at Commons:Commons talk:Naming categories. Commons acts as a central repository of media files to all the wiki projects, and must therefore be available in as many languages as possible to support this work. Thus far this work has been done by hard wiring the language support into the project, by having the same page reproduced multiple times in as many different languages as possible. This has produced resentment when one language is not as well supported as another, and the project becoming frustrated with providing multi-lingual support with regards to category names and categorisation.

This proposal comes about with the realisation that there is no need to individually translate each page and create a copy of that page in each language. With a page translation tool such is available with the likes of Google Language Tools the effort could be directed at creating a "linguistic filter" through which the project could be viewed. A user would choose which language they wish to see the project through and the filter would translate the project into that language.

Key Questions

  • Would not this proposal simply duplicate what is freely available elsewhere?
Initially yes, however sites such as google limit themselves to languages which they see as commercially viable, many African, Indian, classical dead or artifical languages are not catered for. If succesful such a tool will make not only make all wikipedia articles available to everyone, but potentially all websites.
  • How accurate would the tool be?
Poor at first. However just as an wikipedia article has the potential to improve with each reader, the tool can be taught to do its work better with each user. Google language tools has a suggest a better translation option, a very wiki attitude to the world. Users of the tools would not just be passive users but be positively encouraged to teach the tool.
  • How would this differ from available bot translation?
See above, there would be potentially many more language-pairs then any of the free translation sites, the inclusion of users in teaching the bot has the potential of producing more natural and idiomatic translation.
  • Great I like it. How? Where do we start?
One place to start is with the wictionary project, the first thing to do is to create a multi-lingual dictionary by providing interwikis between first words and then phrases that mean the same thing. Already a third party software tool can be used which will lead the user to the definition of the word at Wictionary, such an application could be adopted to instead lead to the definition of the word at a wictionary entry in a different language.
  • How does this differ from proposals to provide a multi-lingual wikipedia
So far as can be told other proposals centre around the creation of a multi lingual wikipedia by bot translating unique wiki articles into other languages as that article is now. As articles are alive and change such a translated wikipedia would soon be left behind. If edited the mulitingual wiki articles would soon diverge into multiple forks as each community has a different take on a subject, and on different takes when multiple articles already exist on a subject which should be the basis of the mass bot translation? This proposal is for just a tool, no version takes precedence, but all will be available depending on the languages supported.
The aims of the two proposals is the same to allow users to use their own languages when using Commons. Proposal:Multilanguage categories aims to do this by modifying the software to support other languages. This proposal attempts to do this by leaving the software alone, and providing a tool which translates the underlying language (normally English) into the users language.
  • If implemented a user would be able to read and navigate a wiki, but how would they edit or contribute to the wiki.
For encyclopaedic articles having access to wiki articles in different languages would allow for the creation of new articles on local wikis. If no existing corresponding article exists on a local wiki, a bot could be instructed to create a page on the local wiki which the user would edit.
For Commons one of the aims of Proposal:Multilanguage categories was to allow users to upload files in their own languages. Just as the wiki is viewed through a layer, the wiki can be loaded through a layer. The uploader would edit to a layer which would then be translated and executed by a bot. An analogy would be like placing an transparency sheet over an image and annotating the image by writing on the sheet. Once the user has finished the transparency sheet would be used to create instuctions for the amending of the image.

Potential Costs

Unknown

References

Community Discussion

Want to work on this proposal?

  1. .. Sign your name here!