Attributed Stream Hypergraphs: temporal modeling of node-attributed high-order interactions

Abstract Recent advances in network science have resulted two distinct research directions aimed at augmenting and enhancing representations for complex networks. The first direction, that of high-order modeling, aims to focus on connectivity between sets nodes rather than pairs, whereas the second one, feature-rich augmentation, incorporates into a all those elements are driven by information which is external structure, like node properties or flow time. This paper proposes novel toolbox, Attributed Stream Hypergraphs (ASHs), unifying both representing, mining, analyzing Applied social analysis, ASHs can characterize phenomena along topological, dynamic attributive elements. Experiments real-world face-to-face online media interactions highlight easily allow analyses, among others, groups’ homophily, nodes’ homophily with respect hyperedges participate, time-respecting paths hyperedges.