Wikidata:Switched On: Working with experts to collate a worldwide database of sexuality topics

Switched On

Working with experts to build a worldwide database of sexuality education

1. Please help integrate the collated data from experts it into Wikidata using Mix n' Match here. Any questions? Ask on the talk page.
2. Suggest, on the talk page ways of displaying and using the data, including how to build something for non Wikimedians to use and contribute to.

Wikipedia and other Wikimedia projects have very large gaps in information. It is estimated that Wikipedia could cover over 100 million topics, currently English Wikipedia, the largest Wikipedia has 6,834,417 articles. There are very large gaps in many topics, for example less than 20% of biographies on English Wikipedia are about women and non binary people. But how do we understand what is missing from Wikipedia?

One way is by working with people with knowledge of the subject to collate databases, reference sources and other resources need to map the missing content. Some parts of Wikimedia are already doing this work, Women in Red is a project which writes about women and non binary people on Wikipedia and uses lists compiled by organisations and crowdsourcing to map who is missing from Wikipedia.

There are a wide range of people with expert knowledge including academics, professionals,, people with (tacit) lived experience, activists, policy makers. Collecting information from a wide range of groups is extremely important to help include as many topics and viewpoints as possible.

Compiled data can also to help others, e.g only 1/5th of experts interviewed in the media worldwide are women, could this change if journalists had an easy to access lists of all experts in specific areas? In turn this would make more reference sources available on women experts, allowing them to be written about on Wikipedia.

But how do we collect this information? Traditionally Wikipedia has found it hard to engage with experts to share information on Wikimedia projects, retention rates for in person training workshops are below 1%. Wikidata provides an opportunity separates knowledge sharing from editing skills which are very often an unrealistic barrier to entry. Wikidata allows us to work with experts to quickly collect information on many topics in a way that can be used by Wikimedia projects in 300 languages.

Switched On was a sexuality education conference organised by UNESCO and UNFPA in early 2020 attended by hundreds of experts and policymakers on different aspects of sexuality education. It had a very strong focus on digital education especially social media and was attended by several Youtube creators and other non traditional educators. It aimed to bring together youth, educators, public health professionals and the tech industry. The conference included a workshop to:

1. Collate expert knowledge on a range of topics related to sexuality.
2. Create a model to use experts's knowledge help map the gaps on Wikidata and Wikipedia in a way that doesn't require them to learn to edit Wikipedia or Wikidata.
3. Create the first version of a worldwide database of sexuality education on Wikidata for use and improvement by sexuality health experts.
4. Improve understanding of gaps on Wikipedia and work with Wikipedia communities to fix those gaps.

This section explains the process of collecting the information:

1. Workshop: A workshop to map expert's knowledge on sexuality education in different categories.
2. Documenting: Type up the post it notes, find duplicates and topics in wrong sections.
3. Observations: understanding patterns and themes in the information collected.
4. Matching the data to Wikidata: importing the data collected into Wikidata.

The workshop was a conference session to educate people about Wikipedia's role in sexuality education and to crowd source people's knowledge on the topic. the presentation used in the workshop is available here. The workshop had four stages:

1. An introduction to Wikipedia, describing what Wikipedia is, its audience and how it fits into the Sustainable Development Goals. Wikipedia was used as a catch all term to save time explaining Wikimedia vs Wikipedia.
2. Examples of projects that can be done with Wikipedia from UN agencies, this included sharing text, images, data and supporting Wikipedia by hosting events and promoting initiatives. Also describing the value of collating data using the FindingGLAMs project to create a worldwide database of cultural heritage institutions as an example.
3. A workshop to collate the expert's knowledge, asking people to share their knowledge; people, projects and organisations, books publications and databases, topics, videos, main messages. People were asked to write down answers to each topic on post it notes for several minutes and then everyone came to the front to compile the information on a large sheet of paper. People could see the information growing as it was added. Note: A workshop was run at a UNFPA event in Nairobi, Kenya which was a less developed version of this workshop, the results have been combined ???..
4. The messages were displayed on a table to allow everyone at the conference to contribute their knowledge, around 20% of the information was added outside of the workshop.

Over 400 different pieces of information were contributed during the 30 minute workshop. There were very few repetitions in the information people provided, showing the volume of information that can be collated in this area of knowledge. The most important improvements to make for the next version of the workshop are to explain the value of crowd sourcing the information of the knowledge in the room earlier in the presentation and asking people to write as clearly as possible. Having people write on post it notes is fun and interactive and looks nice, but it took significant time to write them up, this was a good trade off. There were a small number of topics in wrong section or notes which used acronyms which required some research, some of the handwriting was challenging to read.

• In the workshop 400+ ideas and suggestions were created in half an hour by 20 people, this volume of information demonstrates the breadth of content available in this area. The majority of the people attending the conference were from Europe and North America meaning that local knowledge from other areas of the world was not included.
• Many of the things people would like to see as Wikipedia articles appear to be information they would like to use in their work e.g lists of sexuality education apps, content producers etc. This points to the practical value of Wikipedia in people's professional work.
• Many of the topics people suggested are about how two topics relate to one another e.g gender and feminism, Wikidata and Wikipedia's structure normally focuses on one topic, not the relationship between subjects, it is unclear how to work with this.
• Many of the people that are suggested are from countries that are not well covered in the media e.g Fiji, Eswatini, this shows that there is a significant amount of knowledge that may only be available within small communities and local media or may not be documented at all, since Wikipedia and Wikidata rely on referencing its unclear how these topics can be covered.
• Whilst most of the topics raised can be imported into Wikidata the 'main messages' category does not really fit, its unclear what to do with this information.
• Mix n' Match is being used to import the majority of topics into Wikimedia, however this requires specialist knowledge and skills. Whilst the data could be imported into Wikidata manually it would significantly slow down the process. Easier tools or better instructions for Mix n' Match would really help make this easier.

Matching the lists of topics to Wikidata will take a significant amount of time to achieve, please help by matching the terms listed in section 4 using the Mix n' Match catalogue here.

This section explains possible uses of the data collated and next steps

1. Plan for using the information: mapping out how the data can be used
2. Create first version of a database of sexuality education: Create first version of database of sexuality education, what queries to use, how to display the information, how to add new information
3. Other uses: mapping out additional possibilities for using the data
4. Share work done with experts and ask them to contribute more information: how can the experts who contributed the data make use of it and continue to add to it
5. Feedback and refine: getting feedback from the experts and other users of the data to refine it and uses of it

How can we measure the value of what has been collected during this workshop?

• Mapping the gaps in Wikipedia using the information we have gathered.
• Match the topics to Wikipedia articles or sections within Wikipedia articles that should be created.
• Create a database of sexuality terms.
• Import knowledge from databases and other resources identified by experts using Wikidata:Dataset Imports.

A database that is a useful and usable resource for experts people working the field and simple to add additional information to will need several different decisions to be made:

• What queries to use?
• How to display the information?
• How to add new information in a way that is easy for non Wikimedians?
• Document how to contribute to it?
• What additional documentation within Wikidata needs to be created?

• A similar resource to Women in Red's Wikidata red lists for sexuality related topics for wikiprojects to use

Unsure how to share work done with experts yet

Once we have a first version of the database and other products get feedback from their users, both experts and Wikimedians.

These were the topics people wrote down on post it notes as part of the workshops, topics in italics were suggested at the Nairobi event.