Long-range Meta-path Search on Large-scale Heterogeneous Graphs

17 Jul 2023  ·  Chao Li, Zijie Guo, Qiuting He, Hao Xu, Kun He ·

Utilizing long-range dependency, though extensively studied in homogeneous graphs, has not been well investigated on heterogeneous graphs. Addressing this research gap presents two major challenges. The first is to alleviate computational costs while endeavoring to leverage as much effective information as possible in the presence of heterogeneity. The second involves overcoming the well-known over-smoothing issue occurring in various graph neural networks. To this end, we investigate the importance of different meta-paths and introduce an automatic framework for utilizing long-range dependency on heterogeneous graphs, denoted as Long-range Meta-path Search through Progressive Sampling (LMSPS). Specifically, we develop a search space with all meta-paths related to the target node type. By employing a progressive sampling algorithm, LMSPS dynamically shrinks the search space with hop-independent time complexity. Utilizing a sampling evaluation strategy as the guidance, LMSPS conducts a specialized and effective meta-path selection. Subsequently, only effective meta-paths are employed for retraining to reduce costs and overcome the over-smoothing issue. Extensive experiments on various heterogeneous datasets demonstrate that LMSPS discovers effective long-range meta-paths and outperforms the state-of-the-art. Besides, it ranks top-1 on the leaderboards of \texttt{ogbn-mag} in Open Graph Benchmark. Our code is available at https://github.com/JHL-HUST/LDMLP.

PDF Abstract

Datasets


Task Dataset Model Metric Name Metric Value Global Rank Result Benchmark
Node Property Prediction ogbn-mag LMSPS (w/o embs) Test Accuracy 0.5784 ± 0.0022 # 2
Validation Accuracy 0.5951 ± 0.0007 # 3
Number of params 16470044 # 13
Ext. data No # 1
Node Property Prediction ogbn-mag LMSPS(w/o ComplEx embs) Test Accuracy 0.5767 ± 0.0015 # 4
Validation Accuracy 0.5902 ± 0.0016 # 7
Number of params 16470044 # 13
Ext. data No # 1
Node Property Prediction ogbn-mag LDMLP(w/o ComplEx embs) Test Accuracy 0.5739 ± 0.0012 # 7
Validation Accuracy 0.5888 ± 0.0015 # 8
Number of params 13177884 # 15
Ext. data No # 1

Methods


No methods listed for this paper. Add relevant methods here