Reproduce “HeGMN: Heterogeneous Graph Matching Network for Learning Graph Similarity” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Shilong Sang, Ke-Jia Chen, Zheng liu
HeGMN predicts normalized HGED similarity scores instead of computing exact graph edit distance. It encodes relation types and compares graph-level summaries by node type, while masking node-level attention between different types. On sampled ACM1000 data, removing the mask raises MSE from 0.792 to 0.986 and lowers p@10 from 0.797 to 0.654, supporting the mask’s contribution on this benchmark.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.