Graph Self-Supervised Learning: Taxonomy, Frontiers, and Applications

Content

  • Introduction and background (30 min)
  • Taxonomy of graph self-supervised learning (50 min, Ming) (GSSL survey)
    1. Generation-based graph self-supervised learning
    2. Auxiliary property-based graph self-supervised learning
    3. Contrast-based graph self-supervised learning
    4. Hybrid graph self-supervised learning
  • Frontiers of graph self-supervised learning (50 min, Yizhen)
    1. Trustworthy graph self-supervised learning
    2. Efficient graph self-supervised learning
    3. Automatic graph self-supervised learning
  • Applications of graph self-supervised learning (50 min, Yixin)
    1. Recommender system
    2. Anomaly and out-of-distribution detection
    3. Chemistry
    4. Graph structure learning (SLAPS/SUBLIME)