Wiktionary

The Wiktionary project is an open information project on words and language. It's most known for its dictionary content compiled in some 250+ languages.

Impetus for this page

This page is instigated by Eekim, as per xyr msg on Amgine's talk page asking what support the project feels it might need right now.

Support desires

1. WMF Developer time.
   • To review project-generated tools, code, svn commits.
   • To accomplish project-specific development goals, resolve project-specific bugs.
   • Coddling and encouragement of home-grown volunteers who code for the project.
     • Training/mentoring?
     • Some form of recognition?
2. Project-wide entry specification/standardization
   • Probably would require WMF-brokered negotiations, with hard time tables for development, implementation.
   • I have real trouble with this item. One of the beauties of the projects is that each language community can experiment, can tailor the project to suit their needs. I understand the desire for a uniform spec so that one API can scrape data from all of the projects, but there has to be a more flexible approach that will still permit us to get this done. -- ArielGlenn 18:27, 12 January 2010 (UTC)
     • Specification doesn't say anything about presentation. I would think of it as "this is a list of the standard things which might be included, which should be identified structurally." - Amgine
3. Probably related to the above, support for ongoing intra-project communications.
4. Protection/buffering from Wikipedia interventions in the project, for example the two divisive Logo votes on Meta.
   • Did this have to do with Wikipedia at all? I thought not. -- ArielGlenn 18:28, 12 January 2010 (UTC)
   • No, it did not. I initiated the second vote to compensate for lack of consensus on the first one. As far as the first one, non-English speaking Wiktionarians will disagree that the first one was Wikipedia related as well. Bastique 19:22, 12 January 2010 (UTC)
   • Could someone link to this or write up one or two paragraphs explaining what happened? --Eekim 21:49, 12 January 2010 (UTC)
     • Eekim: The current vote includes a less-than-NPOV brief history, but also links to the original (4 stage) vote. I've also written a couple paragraphs with a couple links. - Amgine 07:31, 14 January 2010 (UTC)
5. New structured content wiki software, NOT a modification of Mediawiki.

   I disagree with this. With more projects like wiktionary:WT:EDIT, the current software can work perfectly for Wiktionary. --Yair rand 17:39, 12 January 2010 (UTC)

   Just so you're aware, the existence of WT:EDIT actually rather proves a need for a structured wiki software, rather than the reverse. If you have to go to that much work to get the flat data to work, to 'make it fit' if you will, you might have less effort/cost - and incidentally make your project much more open to newbie contributors - if you have a more specialized software which abstracts the logic, the content, and the presentation from each other. - Amgine 05:02, 14 January 2010 (UTC)

6. Foundation-supported awareness efforts (i.e., for the Foundation to acknowledge our existence in public)
   • Wikipedia got press releases for 500,000, 1 million, and 10 million total article counts; the 500,000, 1 million, and 2 million English article counts; 1,000 English featured articles; and on and on. There has never (?) been a press release about a non-Wikipedia project, even though the French and English Wiktionaries have each surpassed 500,000, 1 million, and 1.5 million articles, and 5 million articles across all languages at some point in the past, with no fanfare or recognition.
     • I had thought we made some mention of 1 million articles of English and French, but I don't see it on the blog post. I am sure that I was aware of it and shared the information (such as Wikizine). However, I was not specifically notified of that milestone (happened to notice it myself and was on the lookout for it) or the others you mention. It's incumbent on the projects to share this information with the communications division of Wikimedia. Blog posts are easy.Bastique 19:31, 12 January 2010 (UTC)
       • So perhaps that's something we need to delineate and find a process for: how/when/if to notify comcom, or whether they should be notified, or if the Foundation should track that - it's a movement roles issue, right? :) ~Philippe 19:37, 12 January 2010 (UTC)
         • For your information: (fr)[[1]] Otourly 11:06, 13 January 2010 (UTC)

Discussion: What does the project need right now?

One of the intangibles which I feel the project is in need of is the recognition that Wiktionary's content has greater potential, and value, than Wikipedia's for ubiquity in reuse. Every browser uses a dictionary; every word processor has a spell checker and many have a thesaurus. Cellphone apps use DICT servers, captcha's randomly select terms, voicemail are transcribed and forwarded as SMS, and on and on. It would be really useful if WMF would recognize the fundamental worth of having the world's largest online lexical collation, and provide relevant support. - Amgine 07:31, 12 January 2010 (UTC)

New user interfaces. Especially one for quick translation of a word between languages. See sanakirja.org for existing implementation using our wiktionaries. It's very popular in Finland, almost a monopoly in its own right. Similar product without ads could become a killer app. --Ras 09:03, 12 January 2010 (UTC)

Note - Mediawiki might not be the best solution for using/viewing our data. We should also serve/process our existing data in new ways instead of only producing more of it and hoping that it's still the best use of our resources. --Ras 09:10, 12 January 2010 (UTC)

For me the things Wiktionary(ies) need most are (much) more publicity, as was mentioned above already, and additional competent editors (I mean people who know how to make a dictionary). The latter can of course also be improved by training existing editors in things like IPA, some linguistic terminology, and basic lexicography. --Thogo 19:25, 12 January 2010 (UTC)

See also Category:Proposals for Wiktionary

A translation product

If I had to talk to the foundation developers, I would first ask for the main thing which is extremely difficult to realize by the contributors: to provide a translation interface, whose results would be dynamic with the Wiktionaries articles. JackPotte 00:50, 14 January 2010 (UTC)

A multilingual common spam score list

Many of our volunteers "looses their time" currently by reverting some spam, however we've decided to automatize the greatest part of this fight through the following list: wikt:fr:Utilisateur:Salebot/Config. As the email blacklist system, it allows a robot to analyse the contributions contents in order to judge if a revert and a talk page message are needed. This vocabulary should be put in common between the different Wiktionaries, and even for all different Wikis by filtering the Wiktionaries syntax paragraph. JackPotte 21:38, 12 January 2010 (UTC)

A multilingual common Wiktionary gadgets summary

After this proposition, we've listed the best of the Wiktionary gadgets on the French speaker one this week (and it's still continuing). Our conclusions must be known by the other languages projects (or at least which gadgets are installed on which Wiktionaries), for example to avoid to warn individually each site like we were forced to do for this capital one 2 months ago. JackPotte 21:51, 12 January 2010 (UTC)

Needs for smaller wikts

I'm on a wiktionary with an extremely small editing community. Results from our project come up first in google but a majority of the articles are stubs (lacking the definition), and we don't have the people power to add tens of thousands of definitions quickly. So, what we need first are editors. We need a means of attracting new editors and keeping them once we have them.

One piece of that may be to make it easy for passers-by to add to a lemma. Right now it is very hard, since our structure is template-based and that is unlikely to change.

Another thing that might draw new editors and help with content is to find a free (and reliable) source of definitions. En wikt had Webster's 1913, which helped to bootstrap many pages. This can't be done for my language's wiki because all dictionaries out of copyright will be written in an obsolete form of the language. But it might be possible to enter into an agreement with some dictionary publisher such that they would agree to donate perhaps just the definitions, or perhaps all content for some subset of the language (10,000 frequently used words?). For example, Livepedia (www.livepedia.gr) has some sort of arrangement with a publishing house (see [2] for an example where they mention this). -- ArielGlenn 21:55, 12 January 2010 (UTC)