Re: [Wikimediaindia-l] Outcome of CIS work in Kannada Wikipedia
Pavanaja, Thanks for the explaining the outcome of Tulu Workshops and Christ University partnership. While you take a microscopic view of the recent activity and ask me to be optimistic, I would like to be realistic after taking a macroscopic view on past activities. Please see http://cis-india.org/openness/blog/launch-of-assamese-wikipedia-education-program where active editors going from 20 to 20 over a period of 6 months is called 45% growth. The real state of Assamese Wikipedia now after two years can be seen at http://stats.wikimedia.org/EN/TablesWikipediaAS.htm Here is what Asaf from WMF has to say on this: https://meta.wikimedia.org/wiki/Grants_talk:APG/Proposals/2013-2014_round2/The_Centre_for_Internet_and_Society/Proposal_form#Q2a For a tiny Wikipedia like Assamese, it's possible the temporary editing boost leading to a *doubling* of its size by article count and *tripling*of its size by contents was itself the seed of future growth, as the bootstrapping of a Wikipedia is also slow and not self-sustaining work, until that moment when a virtuous cycle kicks in and the usefulness of the resource begins attracting new editors organically. We have perhaps not reached that moment with Assamese, and as you point out, the program is implicitly judged to be less valuable than other opportunities and has thus been discontinued. If this is the case of Assamese Wikipedia which is already out of incubator and that once had a very small but dedicated community, then what is CIS doing working with projects in incubator? Even after CIS working for a year on Konkani Wikipedia, it is not out of incubator. But, it seems you have started the Tulu plan even before the FDC grant is approved. Your FDC proposal staff assessment also notes as follows: https://meta.wikimedia.org/wiki/Grants:APG/Proposals/2013-2014_round2/The_Centre_for_Internet_and_Society/Staff_proposal_assessment CIS’s strategy for its stand-alone projects may not be the most effective for the language communities each project is targeting, given that projects other than Wikipedia (for example, Wikisource or Wiktionary) may be more effective entry-points for working with language communities like Tulu or Santali. To quote Asaf from WMF again: The *sine qua non* of most programs is a core of self-motivating active editors... Where that core doesn't exist, it's very hard to deploy any other type of program... Tulu has a population around 2 million speakers. http://en.wikipedia.org/wiki/Tulu_language You can find a realistic estimate of Editors per million speakers here: http://stats.wikimedia.org/EN/Sitemap.htm For the Indian landscape, Malayalam has around 3 editors (those who make 5+ edits every month) per million and it is the highest (you need to ignore the highly extrapolated value for Sanskrit owing to its tiny population and institutional support). It goes down until 0.2 editors per million for Hindi. To put it in plain words, for every 50 lakh people speaking Hindi, we can hope to get 1 editor making 5+ edits. This trend has been consistent over the years and I don't expect drastic change occurring in the near future unless there is a huge change in socio-economic scenarios. If you look for languages similar to Tulu, Nepal Bhasa comes close. You can check the activity for their Wikipedia at http://stats.wikimedia.org/EN/TablesWikipediaNEW.htm (to be continued.. ) :) Ravi ___ Wikimediaindia-l mailing list Wikimediaindia-l@lists.wikimedia.org To unsubscribe from the list / change mailing preferences visit https://lists.wikimedia.org/mailman/listinfo/wikimediaindia-l
Re: [Wikimediaindia-l] Outcome of CIS work in Kannada Wikipedia
(Continued from http://lists.wikimedia.org/pipermail/wikimediaindia-l/2014-May/011389.html ) Hang on ! While all these stats highlight the people who make 5+ edits, the real strength of any Wikipedia is people who make 100+, 1000+, 2500+ edits per month. Only those Wikipedias which manage to nurture a continuous supply of such editors can progress while keeping in balance editors who go inactive periodically or forever. Let's take the case of Tamil Wikipedia. (Everyone should excuse me quoting Tamil Wikipedia often. It is my home Wikipedia and where have I exclusively worked for many years. I do watch how other Wikipedias work and will try my best to showcase case studies from as many Wikipedias possible. If you know better case studies, please feel free to add to the discussion) Please check the page https://ta.wikipedia.org/wiki/WP:cms Since February 2013, we are keeping track of people who make 100+, 250+ and 1000+ edits every month. So far, in the last 15 months alone, we have 100+ edits - 30 contributors 250+ edits - 41 contributors 1000+ editors - 13 contributors Many of these people hit these milestones often. With so many in the pool, the most number of contributors who have made 100+ in a single month is 34 which happened in January 2014. http://stats.wikimedia.org/EN/TablesWikipediaTA.htm The number of new articles created, daily edits all depend on this most active editor count. For example, check the distribution of article edits in following three languages http://stats.wikimedia.org/EN/TablesWikipediaDE.htm#editdistribution http://stats.wikimedia.org/EN/TablesWikipediaEN.htm#editdistribution http://stats.wikimedia.org/EN/TablesWikipediaML.htm#editdistribution http://stats.wikimedia.org/EN/TablesWikipediaTA.htm#editdistribution Almost universally, the trend of 1-2% editors making 80% of edits is prevalent. And let us see the kind of effort these people put in. During May 2012, just for fun, I asked my fellow Tamil Wikipedians this simple question: https://ta.wikipedia.org/s/qbz How many hours per week do you spend on activities related to Tamil Wikipedia? (Includes time to collect references) 14 people answered and they collectively spent 294 working hours per week. Extrapolating for the stats that month, we found that we spend at least 400+ working hours every week collectively improving Tamil Wikipedia. That is equivalent to employing 10 full time staff just to keep editing Tamil Wikipedia ! And after all this we don't even feel we are anywhere near a very useful and comprehensive encyclopedia. There is lot of work to be done yet. ** CIS is liberal in throwing lot of jargons at the community. Say for example, Needs assessment, Coordination problem, etc., For a difference, I would like to ask them: * Have you ever done a feasibility study for the chances of having a high quality Wikipedia led by active community before engaging in language community plan? Some of the key factors that determined the selection of languages areas as listed at https://meta.wikimedia.org/wiki/India_Access_To_Knowledge/Draft_Work_plan_July_2014_-_June_2015#Language_Area_Work_Plans don't talk about this. That is why I question the sanity of engaging with incubator projects like Tulu. * Are your plans creative enough, big enough to capture and nurture the most active contributors? Without such contributors all efforts will remain one-off and inorganic. That is why I question the effectiveness of paid physical outreach you engage so much. Physical outreach by volunteers can be equally ineffective. But it serves other purposes like community leadership building, brand building etc., and not direct editor recruitment. * What is the lifetime value of one such contributor? In other words, is your spending worth the results you produce? (to be continued.. ) :) Ravi ___ Wikimediaindia-l mailing list Wikimediaindia-l@lists.wikimedia.org To unsubscribe from the list / change mailing preferences visit https://lists.wikimedia.org/mailman/listinfo/wikimediaindia-l
Re: [Wikimediaindia-l] Outcome of CIS work in Kannada Wikipedia
(Continued from http://lists.wikimedia.org/pipermail/wikimediaindia-l/2014-May/011390.html ) OK, where were we :) Ah yes, I was asking about the outcome of CIS's work in general in Kannada Wikipedia. I wanted to hear the story from the community as most of the reports by CIS are unilateral. So, CIS or its partners are under no obligation to answer here. But, as Pavanaja had mentioned that he had answered my questions, I would like to clarify that some of my concerns remain unanswered. * What is the direct impact of numerous Kannada outreach sessions done? As tabulated at https://meta.wikimedia.org/wiki/India_Access_To_Knowledge/Draft_Work_plan_July_2014_-_June_2015/Kannada#Tangible_Impact you have even exceeded the dream target for number of outreach sessions even 6 months in advance ! I wonder why this is even set as a target when it is documented to have negligible value by WMF. Please also go through Wikimedia evaluation portal https://meta.wikimedia.org/wiki/Programs:Evaluation_portal where more such programs are evaluated and best practices are shared. * Regarding Christ University partnership, I would like to have few clarifications. In this mail thread, you have written as follows: //I have personally interacted with many students. There are quite a good number of students who are passionate about enriching Kannada language and bringing latest knowledge into Kannada by way of adding articles to Kannada Wikipedia. Nurturing those students will help not only Kannada Wikipedia but also Kannada language and people in general. Again, I would emphasis, let us be optimistic. // Whereas at https://meta.wikimedia.org/wiki/India_Access_To_Knowledge/Draft_Work_plan_July_2014_-_June_2015/Kannada#Tangible_Impact it is documented as follows: Retaining these as active editors will be a difficult task. Writing article on Wikipedia was a compulsory activity for the students. Only a small percentage of these students have realised the importance of writing on Wikipedia and promised to continue editing Wikipedia. Both are official statements from CIS and are contradicting each other. At least, the optimism shared :) In this context, I have few issues: * Enrolling students compulsorily is a problem that is documented already https://en.wikipedia.org/wiki/Wikipedia:India_Education_Program/Analysis/Independent_Report_from_Tory_Read#The_Foundation_Makes_the_Wikipedia_Assignment_Mandatory * All the upward pointing graphs at https://meta.wikimedia.org/wiki/India_Access_To_Knowledge/Draft_Work_plan_July_2014_-_June_2015/Kannada#Tangible_Impact can mislead the FDC as they don't take into account the project returning to natural growth rates afterwards. * Similar spike and fall is also seen at Hindi Wikipedia http://stats.wikimedia.org/EN/TablesWikipediaHI.htm I assume this could be an effect of the Christ University partnership too as http://cis-india.org/openness/blog/cis-a2k-mou-christ-university says 800 students studying Hindi are also enrolled in this project. Is the Hindi community kept in loop about this effort? Please note that engaging the local community in planning is a recommended best practice as explained at https://en.wikipedia.org/wiki/Wikipedia:India_Education_Program/Analysis/Independent_Report_from_Tory_Read#Engage_the_Global_Wikipedia_Community_as_a_Partner_in_Planning Where it also notes A pilot with more than 1,000 students will result in an unmanageable amount of new material for a finite editor community to monitor If 1000 students is a problem for the gigantic English community, how is the relatively smaller Hindi Wikipedia handling 800 students? How is Kannada Wikipedia handling 600 students? This is especially important as unlike Pune education program, we don't have campus ambassadors, online ambassadors and the entire work load is on the community or the professors (if they are helping). * Is there any data available regarding the quality of content added like the one available at https://en.wikipedia.org/wiki/Wikipedia:India_Education_Program/Analysis/Independent_Report_from_Tory_Read#What_the_Data_Say Given that most of the Indic Wiki communities are still reeling from the work load of Google Translation project, how equipped are the communities to engage these students, clean up and nurture them in to active contributors? * Is CIS engaging in this kind of high volume education program because its previous outings with voluntary enrollment of students did not produce sufficient results? For example, Here are the reports about Assamese and Odia Wikipedia education programs: http://cis-india.org/openness/blog/odia-wikipedia-education-program-iimc-dhenkanal(12 stubs out of a 3 month program) http://cis-india.org/openness/blog/launch-of-assamese-wikipedia-education-program(No clue about results) And that's all for now ! Ravi ___ Wikimediaindia-l mailing list Wikimediaindia-l@lists.wikimedia.org To
[Wikimediaindia-l] How did the Tamil wikipedia achieve quality organic growth - case study?
Dear Ravi, I am impressed by the statistics you have mentioned in your previous mail on the Tamil team in Wikipedia. Can you please share the Tamil team's approach as a case study. I am sure that many of us here will benefit from it. Thank you -- Vikram Vincent ___ Wikimediaindia-l mailing list Wikimediaindia-l@lists.wikimedia.org To unsubscribe from the list / change mailing preferences visit https://lists.wikimedia.org/mailman/listinfo/wikimediaindia-l