Women Occupation Drive

LINK : Women Occupation Drive

About the Women Occupation Drive

Concept :

Let's improve WikiData ! And African languages ! And women representation !

What was the drive about

WikiData biography entries usually refer to the professional occupation of the person. However, in many cases, the only label available is the default form. The fist goal of the drive was to make sure that the top occupations labels recorded in WikiData be available in their male/female/default status. Also, the label is usually available in a limited number of languages (English, German, French...). So the second goal of the drive was to push for the translation of those labels in the languages spoken in Africa. This drive was focusing on Cinema to support the w:en:Wikipedia:WikiProject AfroCine/Months of African Cinema project.

When

October and November 2018

Where

w:en:Wikipedia:WikiProject Wiki Loves Women/Occupations

Who

This is a Wiki Loves Women project organized by user:anthere, with the help of many and in coordination with w:en:Wikipedia:WikiProject AfroCine/Months of African Cinema.

Outcome

• Igbo ig : 46 labels added [4]
• Swahili sw : 35 labels [5]
• Yoruba yo : 9 labels added [6]
• Luganda lg : 5 labels added [7]
• Hausa ha : 4 labels added [8]
• Shona sn : no modifications
• Amharic am : no modifications
• Lingala ln : no modifications
• Afrikaans af : no modifications [9]

For info

• French fr: a few improvements [10]
• Arabic ar : some labels updated [11]
• Portuguese pt : some modifications (do not seem legit...) [12]

Process

• I had in mind to start working on Wikidata to better understand how it worked and how we could use it as a lever for participation, typically to be used during introductory sessions to Wikidata during edit-a-thons/training sessions. But I wanted to avoid us doing work that a bot could perfectly do very well (or even better than us). After several hours discussion with Harmonia, I figured out that the one thing we could better than bot, as human beings, was to work on labels in African languages. So I decided to go ahead with that approach where I could present a table with labels in several African languages, to show clearly the gaps with labels missing.

It so happens that my drive was at the same time than the African Cinema months, so I decided to focus on Cinema related professional occupations. I went digging to identify a list of all occupations entries in WikiData (that was a little bit time consuming part, I downloaded the entire list of occupations, removed the least used ones, and selected the ones related to cinema). Then, with the help of some people on Wikidata, got a query that could run with Listeria bot to automatically update a table with a list of entries. The table displays labels in several languages (some handpicked by myself, some suggested by community members), for generic label, women form label and male form label. The source to help translate was English (perhaps I should have used both French and English). The central page for the contest was published on the English wikipedia. Additions and modifications should be done on WikiData, but the changes are reflected on the wikipedia page thanks to Listeria bot (this provides a complete transparency with regards to changes over time). The page has been kept when the drive was over and the bot maintained, so the page can still be used as a nudge if needed.

the drive was communicated through wikipedia pages, facebook twitter, either from Wiki Loves Women handle or from the African Cinema months handle. And naturally, it was communicated on the African Cinema drive page.

Lessons learned

• It was a bit time consuming to figure out which type of community input would be the most interesting to the Wikidata community. But lesson learned... creating, updating, translating labels is the way to go ;)
• It was a bit time consuming to set-up the system (the wiki page with the table and the bot), but now that it is done, the concept can be easily and quickly extended to other professional occupations (it requires 1) identification of occupations entries Q value in Wikidata and 2) duplication of the table with the new Q entries).
• The suggestion to use LabelLister was initially made. But it revealed a bit difficult to explain users how to switch on LabelLister on their wikidata account. So for the sake of clarity, it was removed
• Outcome is good... but not big. Limited impact unfortunately. Even though we have some community members speaking some of the languages listed. Actually... even though some community members asked me to set up their language and then never show up to fill-in the labels :) But as the page is set-up, it can still be used as a nudge during a training session as part of discovering the interest of WikiData. Still... this is mostly likely to be useful as an ongoing effort, rather than a time limited drive
• some users did not read the instructions (in spite of those being clear and short) and updated the table on wikipedia rather than adding info on wikidata. I am not sure how clearer the instructions may be.
• some labels turned out to be super complicated to translate in some languages... in short... there is no word in some languages and the word usually used is English (very specialized occupations such as ... dub actor)

Future

I do intend to expand it in other occupations areas in the future. It can be either an add-on to other contests focusing on a certain occupation area given the opportunities. Or it may be simply a nudge table.

Context

• 
  Gender Usage per occupation. Screenshot from [1]
• 
  Gender Gap by Occupation in Uganda. Screenshot from [2]
• 
  Gender Gap by Occupation in Tanzania. Screenshot from [3]

Supported by

Wiki Loves Women focuses on bridging two significant gaps on Wikimedia projects – women and Africa – both in terms of content about these subjects and in terms of participation by people from these groups. The project is designed to leverage Wikipedia’s role as a global repository for the dissemination of information to achieve accessible and fair online representation of notable women in countries in Africa. It encourages the contribution of existing researched and verified information to Wikipedia with the intent of redressing the systemic bias online about women. The donated data and content specifically focus on women’s contribution to the political, economic, scientific, cultural and heritage landscape, as well as the current socio-political status of women, in each country. The project will achieve its aims by working with existing gender equality-focused civil society organisations to release their intelligence onto Wikipedia and by working with established Women’s groups and existing Wikipedia Volunteer groups to disseminate this information among the Wikimedia projects. In addition, the project encourage the activation and support of new and existing editors (both female and gender-sensitised male Wikipedians) in the focus countries (Tanzania, Uganda, Côte d'Ivoire, Cameroon, Ghana, and Nigeria).

See also

• https://www.lehir.net/a-tool-to-estimate-gender-gap-on-wikidata-and-wikipedia/
• https://denelezh.dicare.org/gender-gap.php
• http://wdcm.wmflabs.org/WDCM_BiasesDashboard/