Graph Neural Networks for Graphs With Heterophily: A Survey

Recent years have witnessed fast developments of graph neural networks (GNNs) that have benefited myriad graph analytic tasks and applications. Most GNNs rely on the homophily assumption that nodes belonging to the same class are more likely to be connected. However, as a ubiquitous graph property in numerous real-world scenarios, heterophily, i.e., nodes with different labels tend to be linked, significantly limits the performance of tailor-made homophilic GNNs. Hence, GNNs for heterophilic graphs are gaining increasing research attention to enhance graph learning with heterophily. In this paper, we provide a comprehensive review of GNNs for heterophilic graphs. Specifically, we propose a systematic taxonomy…

• NS
  National Science Foundation

  Awards: 2346158, 2106758, III-2106758, POSE-2346158
• CU
  City University of Hong Kong
• NN
  National Natural Science Foundation of China

  Awards: III-2106758, POSE-2346158, 11301224, 62536006