Who Cited It: An AI Research Citation Graph

Field notes

The project fixes its scope to the 3,000 most-cited works in OpenAlex’s Artificial Intelligence subfield and connects papers, authors, institutions, citations, and collaborations. Its methodology treats bibliographic identity as uncertain data: author records receive high, medium, or low confidence labels, with 34% marked low confidence, while 101 paper records are flagged as internally contradictory. The homepage also identifies the corpus definition as a diffable corpus.json file rather than an invisible selection rule.