Integrating heterogeneous structures and community semantics for unsupervised community detection in heterogeneous networks

Yan Zhao, Weimin Li, Fangfang Liu, Jingchao Wang, Alex Munyole Luvembe

Abstract: Highlights•HAESF is an unsupervised community detection framework for heterogeneous networks.•HAE uses two-layer attention and HICG to achieve unsupervised detection.•SF uses semantic information to constrain community detection.•Performance is improved by judgmental loss and segmentation optimizations.