Growing Local Language Content on Wikipedia (Project Tiger 2.0)/Report

From Meta, a Wikimedia project coordination wiki
Project Tiger 2.0 Report


Background

Wikipedia, being a multilingual project, as of August 2020, is available in 312 languages out of which 24 are Indian languages.[1] Indian languages that have the most speakers such as Bengali, Hindi, Punjabi, Tamil, Telugu also have more than 60,000 articles each in their respective Wikipedias. In the last five years, Santali and Tulu Wikipedias went live as well. Thousands of articles in these Indian language Wikipedias are created and improved by many volunteer Wikipedians who spend their valuable time on the growth of Wikipedia and free knowledge in their native language.

This is extremely important to support and strengthen the hard-working Wikimedians so that they can continue contributing to their projects without obstacles as much as possible. Wikimedia Foundation (WMF) and WMF's partner in India Access to Knowledge (A2K) conduct and support a series of events and activities in India such as edit-a-thons, photowalks, Wiki-workshops, equipment support etc. To support and bridge the knowledge gap present online in regional languages, Wikimedia Foundation and Google collaboratively started the Project Tiger initiative in India in 2018. A2K has been the implementation partner for this project in India.

Project Tiger is divided into two phases:

  • Hardware donation: Hardware donation or Hardware support is the first phase of the Project Tiger initiative. Its purpose is to support deserving volunteers in contributing better to Wikimedia projects by providing laptops and/or internet support for 6 months. Before providing the support, criteria are set up so that the most deserving volunteers receive it.
  • Article-writing contest: The second phase, which is not directly linked with the first phase, is essentially a 3-month long edit-a-thon where all the communities participate collaboratively and create/develop articles from the article list given by Google on an array of distributed topics.

The high-level metric for evaluating the impact of the project is ultimately the quality and number of articles created as a result of this project. Project Tiger 2018 witnessed a lot of enthusiasm from participants; over 12 communities took part in the edit-a-thon and created/developed 4,466 articles. Punjabi Wikipedia community won the contest, followed by the Tamil Wikipedia community. An extensive 3-day workshop was organised as a community prize for the top two winner communities in December at Amritsar, Punjab.

Commencement of Project Tiger 2.0

Idea and concept[edit]

A majority of the content present online regarding any topic is available mostly in the English language, thereby disappointing users who prefer to read in or are able to use only their native languages fluently. When Google's research team analysed the results of the most searched yet missing topics on the internet in Indian languages, they decided to partner with the Wikimedia Foundation owing to the fact that Wikipedia is one of the most popular websites available in many languages since several years. They later started a pilot program called Project Tiger, also known as Supporting Indian Language Wikipedias in collaboration with the CIS-A2K to Knowledge in 2017, to support Indic Wikimedians in generating high-quality content in their local language Wikipedias. This project motivates editors to contribute to local languages and free knowledge. It is also known as “Growing Local Language content On Wikipedia" (GLOW) in its second iteration.

Planning and designing of the project[edit]

Project Tiger 2.0 was started in the second half of 2019. Like the first iteration, Project Tiger 2.0 also had two phases of hardware donation and article-writing contest.

Project Tiger was announced in September 2019 over different mailing lists, village pumps and on social media channels.

With the feedback received from volunteers post Project Tiger 1.0 about article lists, we requested communities to create their own set of article lists that are relevant to their respective Wikipedias. Interested candidates were then requested to sign up as a participant on the contest page. Communities were also encouraged to prepare their local event pages inline with the sample page template of the event which was suggested by the foundation. Before starting the contest, an IRC was conducted on 14 September 2019 for the community members to resolve further doubts about the jury for the writing contest, list of community-generated articles etc.

The contest began on 10 October 2019 and the jury started reviewing the articles from 15 October 2019. In this iteration, fifteen Indic language communities including:[2]

A month later, Western Punjabi (written in Shahmukhi script), another script of Punjabi language, showed the interest to participate in Project Tiger and they started to create articles for two months making the total number of communities sixteen. During the period of the contest, Project Tiger also collaborated with Wikipedia Asian Month in November. Communities created articles based on Asia and also submitted under Project Tiger 2.0 fountain too. Project Tiger 2.0 ended on 11 January 2020.

Changes in Project 2.0[edit]

  • Article size: A minimum of 9,000 bytes was required for an article to be evaluated in the previous iteration. Whereas in this iteration, 6,000 bytes was sufficient for all the languages and 3,000 bytes for Urdu articles.[3]
  • Topic lists: In the last iteration, participants only created articles from the list which was provided by Google. But in Project Tiger 2.0, communities were given a chance to generate their own article lists on topics that are relevant to their respective Wikipedias.
  • Before the contest began, an IRC was conducted for all the interested community members to resolve doubts about the jury, list of community-generated articles etc.
  • An open call for a jury for hardware support was put out in this iteration which didn't happen in Project Tiger 2018.

Hardware Donation - Phase 1[edit]

Applications for hardware support were open from 25 August till 14 September and 126 Wikimedians applied for either laptop or internet stipend support or both. Out of 126 individual requests received, 106 applications were for laptops, and 98 were received for internet stipends. We considered only those applications which fulfill the criteria such as:

  • Wikimedians who have a record of contribution to Wikimedia projects of no less than 3 months and no less than 500 contributions across all projects.
  • Only 500 edits on mainspace as of 1 August 2019.
  • To promote diversity and bridge the gender gap, 25% of total laptops and internet stipends were given to women Wikimedians.
  • Wikimedians who already received support earlier from the hardware donation program were not eligible to apply.

In order to help in select the applications and pick 50 laptop recipients and 100 internet stipend recipients, a call for a jury was put out and 8 Wikimedians showed interest (one via email and 7 via meta page). Of these 8 volunteers, 6 members agreed to vet the applications and were declared the jury. After a month of reviewing and analysing the applications based on online contributions till 25 August 2019, offline/outreach contributions (involvement or organizing), objectives (as provided in their applications), future activities, and endorsements, 50 Wikimedians received laptops and 79 received internet stipend for 6 months. As a token of appreciation, the jury members were each given a Project Tiger customized pen and diary. The recipients got their laptops within November or the first week of December whereas the internet stipend process took time till April-May.

Article Writing Contest - Phase 2[edit]

The Project Tiger 2.0 writing contest that began on 10 October 2019, closed on 11 January 2020.[4] The contest ran for three whole months and all the 15 communities participated enthusiastically. Due to the huge number of articles submitted during the contest, the jury took 2 more weeks to evaluate them. After cross-checking fountain tools of all the languages with the help of WMF, the monthly winners and the overall winners of the contest were published on 10 February 2020 via mailing lists, social media pages and on Meta-Wiki pages. A total of 12,406 articles were accepted by the jury and published under the project. Out of the total 331 editors who took part in the writing contest, 68 Wikimedians have received either a laptop or internet stipend support or both, as part of the hardware donation program of Project Tiger 2.0. Top 3 contributors over three months and 40 writing contest jury members across all 15 communities, were sent Amazon Gift vouchers to recognize their hard work and valuable time dedicated to the project.

Offline edit-a-thons

Offline edit-a-thons or on-ground edit-a-thons are generally those Wiki-events where Wikimedians gather in person at one place to edit. The interaction between editors helps in learning better while contributing to WikiProjects and also helps in bringing new editors to any Wikimedia project. As part of Project Tiger 2.0, communities were encouraged to organise edit-a-thons and interested Wikimedians submitted their requests here.
A total of 7 edit-a-thons were organised, the details of which are as follows:

No Community Event name Number of participants Number of articles created/expanded Event page
1 Tamil Wikipedia Workshop and editathon 13 6 Tamil Wikipedia editathon 1
2 Tamil Wikipedia Workshop and editathon 15 3 Tamil Wikipedia editathon 2
3 Telugu Wikipedia (VVIT WikiConnect) Workshop and editathon 7 11 First editathon by VVIT WikiConnect under PT2.0
4 Telugu Wikipedia (VVIT WikiConnect) Workshop and 24 hour editathon 37 171 Project Tiger 2.0 24-hour edit-a-thon
5 Tamil Wikipedia Workshop and editathon 25 2 Tamil Wikipedia editathon 3
6 Punjabi Wikipedia Workshop and editathon 14 20 Project Tiger 2.0 Patiala Meetup
7 Kannada Wikipedia (Alva's Wikipedia Student's Association) Workshop and editathon 7 76 Alva's Wikipedia Project Tiger Edit-a-thon

Stories from the Hardware recipients

Under the Project Tiger 2.0, a total of 126 recipients got hardware support which includes 50 editors who received laptops and 98 who received the internet stipend. As feedback was being collected from editors about the entire contest, recipients of hardware support were requested to share their stories with CIS-A2K, and a few of them got back with the same. Here are a few such stories:

User:Ssgapu22
User:Ssgapu22 is one of the most active contributors who has successfully completed #1000Wikidays on Odia Wikipedia. He was a Chromebook recipient in Project Tiger 2018 and internet support recipient in this iteration. He has served as a jury member for the Odia community during Project Tiger 2.0. He says :

"During the 1st time, I got the Chromebook, so this time I had applied for the internet. Nitesh Gill and Suswetha Kolluru, the Project Tiger Co-ordinators from CIS-A2K had helped me a lot from the beginning.

Before starting this project, I had participated in Project Tiger Community Consultation at Chennai in March 2019, where we had planned for this edit-a-thon. This edit-a-thon started from the 10th of October in 2019 and ended on 10th January 2020. This time, we got the chance to choose the articles for this edit-a-thons from which we had created just 155 articles. Dr Subas Chadran Rout had created 76 articles, Aliva Sahoo had created 48, Harihar Satpathy had created 24 articles, and I had created just 6. In those days I had an issue with the company where I was working, so couldn't create many articles, but overall we had improved from the previous edit-a-thon. In 2018, I was in Bangalore and had participated in an offline edit-a-thon at CIS-A2K office, Bangalore, where I had created 4 or 5 articles in one day, but at this time, we couldn't conduct any offline edit-a-thon, which might help to create more articles. The overall experience is good and for this edit-a-thon, some new and good articles were created on Odia Wikipedia. I couldn't participate more in this edit-a-thon, but the internet support is still on, which is definitely helping me to create articles every day for the #100wikidays.

I must thank all the co-ordinators from Wikimedia Foundation, CIS-A2K, local Jury members, all the participants, and the Google team for this project."


User:Dugal harpreet
User:Dugal harpreet from Punjabi Wikipedia, one of the active contributors in the Punjabi community who usually edits Punjabi Wikisource notes that:

"ਪ੍ਰੋਜੈਕਟ ਟਾਈਗਰ ਅਧੀਨ ਮਿਲੇ ਲੈਪਟਾਪ ਕਾਰਨ ਮੈਨੂੰ ਬਹੁਤ ਸਹਾਇਤਾ ਮਿਲੀ ਹੈ। ਇਸ ਤੋਂ ਪਹਿਲਾਂ ਮੈਂ ਆਪਣੇ ਫੋਨ ਤੋਂ ਹੀ ਵਿਕੀਸਰੋਤ ਤੇ ਕੰਮ ਕਰਦੀ ਸੀ। ਜਿਸ ਨਾਲ ਮੈਨੂੰ ਬਹੁਤ ਮੁਸ਼ਕਿਲਾਂ ਦਾ ਸਾਹਮਣਾ ਕਰਨਾ ਪੈਂਦਾ ਸੀ। ਫੋਨ ਤੋਂ ਜਿਆਦਾ ਸਮੇਂ ਤੱਕ ਕੰਮ ਵੀ ਨਹੀਂ ਹੁੰਦਾ ਸੀ। ਲੈਪਟਾਪ ਮਿਲਣ ਤੋਂ ਪਹਿਲਾਂ ਮੈਂ ਸਿਰਫ਼ ਵਿਕੀਸਰੋਤ ਤੇ ਹੀ ਕੰਮ ਕਰਦੀ ਸੀ, ਕਿਉਂਕੀ ਫੋਨ ਤੋਂ ਵਿਕੀਪੀਡੀਆ ਤੇ ਆਰਟੀਕਲ ਬਣਾਉਣ ਵਿੱਚ ਮੁਸ਼ਕਿਲ ਆਉਂਦੀ ਸੀ। ਲੈਪਟਾਪ ਮਿਲਣ ਕਾਰਨ ਮੈਂ ਵਿਕੀਪੀਡੀਆ ਤੇ ਵੀ ਕੰਮ ਕਰਨਾ ਸ਼ੁਰੂ ਕੀਤਾ ਹੈ। ਫੋਨ ਤੋਂ ਕਿਤਾਬ ਪਰੂਫਰੀਡ ਕਰਦੇ ਸਮੇਂ ਕਰੋਪ ਈਮੇਜ਼ ਵਾਲੀ ਟੈਂਪਲੇਟ ਕੰਮ ਨਹੀਂ ਕਰਦੀ ਸੀ। ਜਿਸ ਕਾਰਨ ਮੈਨੂੰ ਫੋਟੋ ਵਾਲੇ ਪੇਜ਼ ਬਿਨ੍ਹਾਂ ਪਰੂਫਰੀਡ ਕੀਤੇ ਛੱਡਣੇ ਪੈਂਦੇ ਸੀ। ਪਰ ਹੁਣ ਮੈਨੂੰ ਵਿਕੀਸਰੋਤ ਤੇ ਵਿਕੀਪੀਡੀਆ ਤੇ ਪਰੂਫਰੀਡ ਕਰਨ ਵਿੱਚ ਕੋਈ ਦਿੱਕਤ ਨਹੀਂ ਆਉਂਦੀ। ਪਹਿਲਾਂ ਜਦੋਂ ਮੈਂ ਜਦੋ ਜਾਂ ਦੋ ਤੋਂ ਵੱਧ ਬੰਦਿਆਂ ਨੂੰ ਵਿਕੀਪੀਡੀਆ ਤੇ ਕੰਮ ਕਰਨਾ ਦੱਸਦੀ ਸੀ, ਤਾਂ ਫੋਨ ਤੋਂ ਸਿਖਾਉਣਾ ਔਖਾ ਹੋ ਜਾਂਦਾ ਸੀ ਕਿਉਂਕਿ ਫੋਨ ਤੋਂ ਜਿਆਦਾਤਰ ਲੋਕ ਵਿਕੀਪੀਡੀਆ ਬਾਰੇ ਸਿੱਖਣ ਵਿੱਚ ਔਖ ਮਹਿਸੂਸ ਕਰਦੇ ਹਨ। ਇਸ ਲਈ ਹੁਣ ਮੈਂ ਲੈਪਟਾਪ ਤੋਂ ਲੋਕਾਂ ਨੂੰ ਵਿਕੀਪੀਡੀਆ ਤੇ ਕੰਮ ਕਰਨਾ ਆਸਾਨੀ ਨਾਲ ਸਿਖਾ ਸਕਦੀ ਹਾਂ ਅਤੇ ਚੰਗੀ ਤਰ੍ਹਾਂ ਸਮਝਾ ਵੀ ਸਕਦੀ ਹਾਂ। ਲੈਪਟਾਪ ਦੀ ਮਦਦ ਮਿਲਣ ਕਰਕੇ ਮੈਂ ਇੰਗਲਿਸ਼ ਵਿਕੀਪੀਡੀਆ ਤੇ ਵੀ ਕੰਮ ਕਰਨਾ ਸ਼ੁਰੂ ਕੀਤਾ ਹੈ।" "The laptop I got under Project Tiger helped me a lot. Before that, I used to work on Wikisource on my phone which caused me a lot of trouble. The phone didn't work for long. Before I got the laptop, I only worked on Wikisource, because it was difficult to create articles on Wikipedia from the phone. Having got a laptop, I also started working on Wikipedia. I was not able to work on templates, images, or technical work while proofreading the book over the phone and I usually had to leave the photo pages without proofreading. But now I have no problem to proofread the books as well as edit Wikipedia. In the past, when I teach anyone to work on Wikipedia, it became difficult to explain to others over the phone because most people find it difficult to learn. From when I received a laptop, I can easily show and teach people how to work on Wikipedia and also explain it well. With the help of a laptop, I have also started working on the English Wikipedia."

User:Harshil169

"The laptop which I got is extremely helpful while editing Wikimedia projects. A lot of my work has become easier because previously I used to edit via mobile which was quite difficult. I was able to contribute much better to my home Wiki i.e, Gujarati Wikipedia by creating new pages. I have also uploaded photos to the commons using my laptop. Would like to appreciate CIS for this project."

User:Sumita Roy Dutta and User:Goutam1962

"During Tiger project 2.0, myself and my husband Goutam have both received a Windows 10 based laptop which was a dream come true for both of us and we had applied for net connection and that too was granted. Before that, Goutam was contributing through Redmi 6A mobile and me, with my mobile because our old ubuntu-based laptop was not working. After receiving the laptop, Goutam came first in Hindi Tiger project 2.0 (third month) and with his new laptop, his contribution to Wikidata and English Wikipedia also increased tremendously. He had created 100 articles out of which 87 articles on different peaks of Uttarakhand were accepted and published. Post Project Tiger, I had uploaded images of different railway stations in India, then was busy in updating Wikidata (COVID-19) related data, same on English Wikipedia and Bengali Wikipedia and uploaded COVID related images in Commons. Worked for Wiki Loves Folklore images and articles and category creation. Voluntarily retired persons like us having hardware support like laptop and net connection who have to move from Delhi to Kolkata every six months (to take care of old parents ) was a basic need to hassle-free contribution on Wikipedia. Thanks to CIS for supporting our dream and hope other new Wikipedians will also receive support from CIS and project GLOW will glow forever.

User:Vijay Barot

"I would like to share some personal inputs. I am not using a personal smartphone so for me, internet stipend is very useful and I mostly contribute at late nights due to unavailability of personal computers. My parents were not happy with my unrewarding contribution but when I got a laptop they felt very happy and now I can contribute more freely.

Metrics

331 editors from 15 Indic Wikipedia communities participated in the writing contest. A total of 12,406 articles were created and accepted by the jury when the contest ended. The highest number of articles were created by the Tamil community of 62 participants with 2,924 articles. Second and third places were occupied by the Punjabi community with 1,747 articles followed by the Bengali community with 1,194 articles.

Languages Articles submitted Articles accepted Fountain tool
Tamil 2,959 2,924 Link
Gujarati 202 201 Link
Santali 566 566 Link
Tulu 32 32 Link
Punjabi 1,768 1,747 Link
Kannada 249 199 Link
Sanskrit 19 19 Link
Odia 155 154 Link
Hindi 417 253 Link
Marathi 220 97 Link
Urdu 1,377 1,376 Link
Malayalam 229 229 Link
Telugu 416 24 Link
Bengali 1,460 1,194 Link
Assamese 205 175 Link

Learnings

Group photo of participants of VVIT WikiConnect during Project Tiger 2.0 24hrs Edit-a-thon

After the completion of Project Tiger 2.0 article writing contest and hardware donation process, we conducted a detailed survey to understand the participants' experience during the project. With the support of Wikimedia Foundation, we created feedback forms for the edit-a-thon participants and hardware support recipients. 105 Wikimedians participated in the survey, 78 of them were article writing contest and hardware support recipients, and 27 were jury members of the writing contest. Some of the suggestions we received are:

  • Several Wikimedians suggested conducting detailed online orientation sessions before starting the article-writing contest. The session(s) might be related to Wikipedia article writing "how-to" answers, Wikipedia policies, guidelines, and manual of style or the rules of Project Tiger itself.
  • We have got mixed feedback about the article size. In Project Tiger 2.0 article size was fixed at 6,000 bytes. While many agreed that 6,000 bytes are sufficient for a well-written article, a few believe at least 9,000 bytes are necessary for an informative article on Wikipedia.
  • A few participants informed us that they did not get regular updates related to Project Tiger contest on their talk pages.
  • This year participants also suggested that collaboration of Project Tiger with other contests results in misunderstandings with respect to rules and that it can disrupt the participants' enthusiasm.

Wikipedians noted that during Project Tiger 2.0 prizes were not given in time, which can be improved in the next iteration. We have received suggestions to conduct more regular online sessions, and communication with the communities and also the jury.

Several jury members of the writing contest gave inputs regarding the Fountain tool. They include:

  • Making the 'manage' option available for a few more days after the contest ends, as this option is necessary for evaluation.
  • Fountain tool closing time should match with contest end time i.e, IST.

We'll use our learning to improve the next iteration of Project Tiger or similar event(s) better, and we'll make changes in the program according to the communities' need.

See also

Footnotes

  1. Other than English language
  2. Assamese, Tulu and Santali were the new additions in Project Tiger 2.0
  3. This decision was made based on the feedback received after Project Tiger 2018
  4. There was a delay by one day due to a fountain tool error