Empowering Dual-Level Graph Self-Supervised Pretraining with Motif Discovery
Pengwei Yan, Kaisong Song, Zhuoren Jiang, Yangyang Kang, Tianqianjin Lin, Changlong Sun, Xiaozhong Liu
Abstract
While self-supervised graph pretraining techniques have shown promising results in various domains, their application still experiences challenges of limited topology learning, human knowledge dependency, and incompetent multi-level interactions. To address these issues, we propose a novel solution, Dual-level Graph self-supervised Pretraining with Motif discovery (DGPM), which introduces a unique dual-level pretraining structure that orchestrates node-level and subgraph-level pretext tasks. Unlike prior approaches, DGPM autonomously uncovers significant graph motifs through an edge pooling module, aligning learned motif similarities with graph kernel-based similarities. A cross-matching task enables sophisticated node-motif interactions and novel representation learning. Extensive experiments on 15 datasets validate DGPM's effectiveness and generalizability, outperforming state-of-the-art methods in unsupervised representation learning and transfer learning settings. The autonomously discovered motifs demonstrate the potential of DGPM to enhance robustness and interpretability.
BibTeX
@article{Yan_Song_Jiang_Kang_Lin_Sun_Liu_2024, title={Empowering Dual-Level Graph Self-Supervised Pretraining with Motif Discovery}, volume={38}, url={https://ojs.aaai.org/index.php/AAAI/article/view/28774}, DOI={10.1609/aaai.v38i8.28774}, abstractNote={While self-supervised graph pretraining techniques have shown promising results in various domains, their application still experiences challenges of limited topology learning, human knowledge dependency, and incompetent multi-level interactions. To address these issues, we propose a novel solution, Dual-level Graph self-supervised Pretraining with Motif discovery (DGPM), which introduces a unique dual-level pretraining structure that orchestrates node-level and subgraph-level pretext tasks. Unlike prior approaches, DGPM autonomously uncovers significant graph motifs through an edge pooling module, aligning learned motif similarities with graph kernel-based similarities. A cross-matching task enables sophisticated node-motif interactions and novel representation learning. Extensive experiments on 15 datasets validate DGPM’s effectiveness and generalizability, outperforming state-of-the-art methods in unsupervised representation learning and transfer learning settings. The autonomously discovered motifs demonstrate the potential of DGPM to enhance robustness and interpretability.}, number={8}, journal={Proceedings of the AAAI Conference on Artificial Intelligence}, author={Yan, Pengwei and Song, Kaisong and Jiang, Zhuoren and Kang, Yangyang and Lin, Tianqianjin and Sun, Changlong and Liu, Xiaozhong}, year={2024}, month={Mar.}, pages={9223-9231} }