Research map / Graph learning / Graph self-supervised learning
Type-Aware Anchor Link Prediction across Heterogeneous Networks Based on Graph Attention Network
Xiaoxue Li, Yanmin Shang, Yanan Cao, Yangxi Li, Jianlong Tan, Yanbing Liu
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add type-aware matching to fusion-only alignment (TALPf -> TALP) (ablation) | TALP · Twitter-Foursquare | P@1 (%): 42.66 -> 43.79 | Improved |
| add type-aware matching to fusion-only alignment (TALPf -> TALP) (ablation) | TALP · Aminer-Mag | MAP@30 (%): 66.25 -> 82.39 | Improved |
| add fusion-vector matching to type-aware-only alignment (TALPa -> TALP) (ablation) | TALP · Twitter-Foursquare | MAP@30 (%): 57.68 -> 59.33 | Improved |
| add fusion-vector matching to type-aware-only alignment (TALPa -> TALP) (ablation) | TALP · Aminer-Mag | MAP@30 (%): 79.49 -> 82.39 | Improved |
| use type-aware only instead of fusion only (ablation) | TALP · Aminer-Mag | MAP@30 (%): 66.25 -> 79.49 | Improved |
| replace HAN/PME/HHNE-style fusion embedding with TALP's GAT fusion (TALPf vs PME) | TALPf · Aminer-Mag | P@1 (%): 45.91 -> 50.56 | Improved |
| use TALP instead of strongest baseline PME | TALP · Aminer-Mag | MAP@30 (%): 55.96 -> 82.39 | Improved |
| use TALP instead of strongest baseline PME | TALP · Twitter-Foursquare | MAP@30 (%): 52.42 -> 59.33 | Improved |
| set type-aware weight (1-omega) to 0 (fusion only) (ablation) | TALP · Twitter-Foursquare and Aminer-Mag | P@k / MAP | Hurt |
| increase type-aware weight (1-omega) beyond optimum (ablation) | TALP · Twitter-Foursquare and Aminer-Mag | P@k / MAP | Mixed |
| increase attention heads K beyond 3 (ablation) | TALP · Twitter-Foursquare and Aminer-Mag | P@k / MAP | Hurt |
Closest papers by meaning
- Link Prediction in Multilayer Networks via Cross-Network Embedding (AAAI 2024 · similarity 0.87)
- On the Impact of Feature Heterophily on Link Prediction with Graph Neural Networks (NeurIPS 2024 · similarity 0.87)
- A Topological Perspective on Demystifying GNN-Based Link Prediction Performance (ICLR 2024 · similarity 0.85)
- Revisiting Link Prediction: a data perspective (ICLR 2024 · similarity 0.85)
- Future Link Prediction Without Memory or Aggregation (NeurIPS 2025 · similarity 0.84)
- TAMI: Taming Heterogeneity in Temporal Interactions for Temporal Graph Link Prediction (NeurIPS 2025 · similarity 0.84)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
