Strengthening Wikipedia with candidate citations into academic literature at scale (200M+) using contemporary AI techniques
🎥 Session recording: [https://commons.wikimedia.org/wiki/File:Wikimania_2026_%E2%80%93_Strengthening_Wikipedia_with_candidate_citations_into_academic_literature_at_scale_(200M%2B)_using_contemporary_AI_techniques.webm](https://commons.wikimedia.org/wiki/File:Wikimania_2026_%E2%80%93_Strengthening_Wikipedia_with_candidate_citations_into_academic_literature_at_scale_(200M%2B)_using_contemporary_AI_techniques.webm)
Building on the Internet Archive’s work to protect Wikipedia’s references Turn All References Blue, the Internet Archive Europe set the challenge to use contemporary AI techniques to contribute candidate references from academic articles to Wikipedia editors at scale.
The workshop will demo the initial version of the system and introduce editors to the easy to use system. Archive Sources uses a pipeline of contemporary AI techniques to find academic papers relevant to claims made in Wikipedia articles, from an index of 237 million academic passages. These paper passages may back up existing claims, add missing citations, or surface newer research that updates or nuances what the article currently says. Future work will rank the candidate citation passages to English wikipedia passages in databases.
You must log in to leave a comment.
