Research map / Graph learning / Graph self-supervised learning
Evolutionarily Learning Multi-Aspect Interactions and Influences from Network Structure and Node Content
Songlei Jian, Liang Hu, Longbing Cao, Kai Lu, Hang Gao
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 multi-aspect influence propagation system to MAI-LS (MAI-LS -> MAI-ECS) (ablation) | MAI-LS · Aminer citation node classification, 80% | accuracy: 0.857 -> 0.887 | Improved |
| add multi-aspect influence propagation system to MAI-LS (ablation) | MAI-LS · Aminer citation node classification, 80% | F1-score: 0.802 -> 0.832 | Improved |
| add multi-aspect influence propagation system to MAI-LS (ablation) | MAI-LS · Aminer citation node classification, 20% | accuracy: 0.845 -> 0.873 | Improved |
| add multi-aspect influence propagation system to MAI-LS (ablation) | MAI-LS · Aminer citation node classification, 20% | F1-score: 0.783 -> 0.815 | Improved |
| add influence propagation to MAI-LS (ablation) | MAI-LS · Aminer Coauthor link prediction | nDCG@1: 0.956 -> 0.965 | Improved |
| add influence propagation to MAI-LS (ablation) | MAI-LS · Aminer Citation link prediction | nDCG@1: 0.805 -> 0.820 | Improved |
| add influence propagation to MAI-LS (ablation) | MAI-LS · Aminer Citation link prediction | nDCG@10: 0.889 -> 0.898 | Improved |
| use network structure (Node2vec) instead of node content only (Doc2vec) | embedding input · Aminer citation node classification, 80% | accuracy: 0.658 -> 0.804 | Improved |
| use network structure (Node2vec) instead of node content only (Doc2vec) | embedding input · Aminer Citation link prediction | nDCG@1: 0.385 -> 0.716 | Improved |
| concatenate Node2vec and Doc2vec (NV+DV) vs Node2vec alone | embedding input · Aminer citation node classification, 80% | accuracy: 0.804 -> 0.838 | Improved |
| use NV+DV concatenation instead of CANE with separate objectives | baseline choice · Aminer citation node classification, 80% | accuracy: 0.820 -> 0.838 | Improved |
| use NV+DV concatenation instead of TriDNR | baseline choice · Aminer citation node classification, 80% | accuracy: 0.817 -> 0.838 | Improved |
Closest papers by meaning
- Distilling and Adapting: A Topology-Aware Framework for Zero-Shot Interaction Prediction in Multiplex Biological Networks (ICLR 2026 · similarity 0.87)
- Joint Identification of Network Communities and Semantics via Integrative Modeling of Network Topologies and Node Contents (AAAI 2017 · similarity 0.87)
- TAMI: Taming Heterogeneity in Temporal Interactions for Temporal Graph Link Prediction (NeurIPS 2025 · similarity 0.85)
- On the Impact of Feature Heterophily on Link Prediction with Graph Neural Networks (NeurIPS 2024 · similarity 0.85)
- HeterSEED: Semantics–Structure Decoupling for Heterogeneous Graph Learning under Heterophily (NeurIPS 2026 · similarity 0.84)
- Learning Changes in Graphon Attachment Network Models (ICML 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.
