Research map / Graph learning / Graph self-supervised learning
Multi-graph Fusion Cross-model Contrastive Learning for Recommendation
Shengjun Ma, Yuhai Zhao, Fenglong Ma, Baoyin Liu, Zhengkui Wang, Wen Shan
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 cross-model contrastive learning (CMC) (ablation) | MFCCL · Alibaba-iFashion | recall@20: 0.1094 -> 0.1288 | Improved |
| add cross-model contrastive learning (CMC) (ablation) | MFCCL · Last-FM | recall@20: 0.0883 -> 0.1025 | Improved |
| use virtual relations instead of original KG relations (ablation) | MFCCL · Alibaba-iFashion | ndcg@20: 7.21% relative | Improved |
| use virtual relations instead of original KG relations (ablation) | MFCCL · Last-FM | ndcg@20: 0.0897 -> 0.0978 | Improved |
| add user-interest graph (UiG) (ablation) | MFCCL · Alibaba-iFashion | recall@20: 0.1276 -> 0.1288 | Improved |
| add bipartite graph for user representation (ablation) | MFCCL · Alibaba-iFashion | recall@20: 0.1264 -> 0.1288 | Improved |
| add item-association graph (IaG) (ablation) | MFCCL · Last-FM | recall@20: 0.1012 -> 0.1025 | Improved |
| add KG for item representation (ablation) | MFCCL · Last-FM | recall@20: 0.1012 -> 0.1025 | Improved |
| MFCCL over best baseline | MFCCL · Alibaba-iFashion | recall@20: 0.1196 -> 0.1288 | Improved |
| MFCCL over best baseline | MFCCL · Yelp2018 | ndcg@20: 0.0604 -> 0.0501 | Hurt |
| more virtual relations than moderate value (ablation) | MFCCL · Alibaba-iFashion, Last-FM | recall@20, ndcg@20 | Hurt |
| set CL loss coefficient to 0.0001 (ablation) | MFCCL · three datasets | recall@20 | Improved |
Closest papers by meaning
- Robust Graph Based Social Recommendation Through Contrastive Multi-View Learning (AAAI 2025 · similarity 0.89)
- DivGCL: A Graph Contrastive Learning Model for Diverse Recommendation (AAAI 2025 · similarity 0.88)
- Graph Disentangled Contrastive Learning with Personalized Transfer for Cross-Domain Recommendation (AAAI 2024 · similarity 0.88)
- Review-Enhanced Hierarchical Contrastive Learning for Recommendation (AAAI 2024 · similarity 0.88)
- Spectral-Based Graph Neural Networks for Complementary Item Recommendation (AAAI 2024 · similarity 0.87)
- Semantic Enhanced Heterogeneous Hypergraph Network for Collaborative Filtering (AAAI 2025 · similarity 0.87)
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.
