Research:Item Recommendations for Wikidata Editors
This project develops a recommender system that recommends Wikidata items to editors based on their past editing activities. The work is motivated by Wikidata's quest for more and more engaged editors to keep up with the knowledge graph of growing size and complexity. The aims of the projects: - Create a foundation for understanding how Wikidata items can be recommended to the editors, Examine whether implementing a recommender system helps editors and increases their engagement.
The project will consist of the followings steps: 1- Interview study with Wikidata editors to explore their current practices into how to choose that they work on, and elicit their preferences and needs. 2- Develop the recommendation model accordingly 3- Evaluate the model using offline evaluation (on a dataset) as well as online evaluation with real editors. The link to the interview questions is: https://docs.google.com/document/d/1flMpstjFbzlD_AfMTsKM0Ry985jLPGo4eV6vZHi_hJ4/edit?usp=sharing
Policy, Ethics and Human Subjects Research
The interview study and online evaluation was approved by the local committee of the developers' institution (King's College London), with the registration confirmation reference number, MRSP-20/21-23336