OpenAlex: An open and comprehensive index of scholarly works, citations, authors, and institutions

"OpenAlex is a free and open Scientific Knowledge Graph (SKG).  It contains information describing approximately 230M scholarly works, drawn from both structured (eg: Crossref) and unstructured (eg: institutional repositories, publisher websites) sources, clustered/merged into distinct records, and linked by citations. By parsing work metadata and enriching it with external PID sources (ROR, ORCID, ISSN Network, PubMed, Wikidata, etc), OpenAlex describes and links (approximately) 200M author clusters, 100k institutions, and100k venues (journals and repositories). Using a neural-net classifier, we assign one or more of 50k Wikidata concepts to each work. All source code and ML models are available openly, and data is freely available via a high-performance API, a complete database dump, and a search-engine-style web interface. This talk will describe the construction of OpenAlex, compare it to other SKGs (eg Scopus, MAG), and discuss plans for the future."