math-cocitations

Summary:The math-cocitations dataset is a co-citation hypergraph of highly cited Mathematics papers, constructed from the OpenAlex Snapshot (2024-09-27; https://developers.openalex.org/download/snapshot-format). Nodes and Hyperedges:Nodes represent highly cited Mathematics papers. For each Mathematics subfield, papers were ranked by citation count, and those accounting for the top 10% of cumulative citations were selected as nodes (minimum 100 papers per subfield), yielding 2,972 nodes in total. Hyperedges represent co-citation relationships. For each paper in the OpenAlex corpus (excluding the selected top papers), the subset of top papers it cites was identified. If the same subset appeared in at least 3 citing papers, it was included as a hyperedge, yielding 17,099 hyperedges in total. Each node carries the following attributes: OpenAlex work ID, paper title, publication date, primary topic, subfield, field, domain, and citation count. Basic statistics:- Nodes: 2,972- Hyperedges: 17,099 Source:OpenAlex Snapshot (2024-09-27), https://developers.openalex.org/download/snapshot-format Reference:Kazuki Nakajima, Yuya Sasaki, Takeaki Uno, and Masaki Aida. (2025). Learning Multi-Order Block Structure in Higher-Order Networks. arXiv preprint arXiv:2511.21350.