Fig. 8: Two case studies for hypergraph analysis and hypergraph learning on real-world datasets. | Humanities and Social Sciences Communications

Fig. 8: Two case studies for hypergraph analysis and hypergraph learning on real-world datasets.

From: EasyHypergraph: an open-source software for fast and memory-saving analysis and learning of higher-order networks

Fig. 8: Two case studies for hypergraph analysis and hypergraph learning on real-world datasets.

a–f Hypergraph analysis on House-committees and Senate-committees datasets. a The cumulative frequency of hyperdegree. b The frequency of hyperedge degree. c The frequency of cycle ratio and the percentage of members from Democrat or Republican beyond the average cycle ratio. d The distributions of hypercoreness value and corresponding political party affiliation of members. e The Spearman correlation coefficient between s-closeness and s-betweenness across different interaction sizes. f The cumulative frequency of s-closeness (s = 1) of original hypergraph and generative hypergraph. The generative hypergraph is generated by the Chung-Lu model. g, h Node classification on co-citation Cora and co-citation CiteSeer datasets by training hypergraph learning models.

Back to article page