Research map / Graph learning / Graph self-supervised learning
Revisiting Contrastive Learning in Collaborative Filtering via Parallel Graph Filters
Fang Kai, Yu Zhang, Kaibin Wang, Lei Sang, Yiwen Zhang
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 |
|---|---|---|---|
| replace LightCCF with margin-constrained cosine similarity (LightCSCF) | LightCCF · Amazon-book | Recall@20: 0.0569 -> 0.0594 (4.39%) | Improved |
| replace LightCCF with margin-constrained cosine similarity (LightCSCF) | LightCCF · Tmall | Recall@20: 0.0930 -> 0.0972 (4.52%) | Improved |
| replace LightCCF with margin-constrained cosine similarity (LightCSCF) | LightCCF · Douban-book | Recall@20: 0.1863 -> 0.1899 (1.93%) | Improved |
| replace RCL-based SCCF cosine similarity with margin-constrained similarity | SCCF · Amazon-book | Recall@20: 0.0491 -> 0.0594 (13.14%) | Improved |
| replace RCL-based SCCF cosine similarity with margin-constrained similarity | SCCF · Tmall | Recall@20: 0.0772 -> 0.0972 (13.95%) | Improved |
| replace RCL-based SCCF cosine similarity with margin-constrained similarity | SCCF · Douban-book | Recall@20: 0.1711 -> 0.1899 (10.99%) | Improved |
| replace RCL-based SCCF cosine similarity with margin-constrained similarity | SCCF · Tmall | NDCG@20: 0.058 -> 0.0741 (17.25%) | Improved |
| increase embedding size from 64 to 1024 (ablation) | LightCSCF · Amazon-book, Tmall, Douban-book | Recall and NDCG | Mixed |
| inject noise into interactions (10% to 40%) (ablation) | LightCSCF · Amazon-book, Douban-book | NDCG@20 | Mixed |
| add margin-constrained loss versus LightGCN (3 layers) | LightGCN · Tmall | Recall@20: 0.0711 -> 0.0972 | Improved |
Closest papers by meaning
- DivGCL: A Graph Contrastive Learning Model for Diverse Recommendation (AAAI 2025 · similarity 0.89)
- Hgformer: Hyperbolic Graph Transformer for Collaborative Filtering (ICML 2025 · similarity 0.88)
- Robust Graph Based Social Recommendation Through Contrastive Multi-View Learning (AAAI 2025 · similarity 0.87)
- Temporal Graph Contrastive Learning for Sequential Recommendation (AAAI 2024 · similarity 0.87)
- Beyond Graph Convolution: Multimodal Recommendation with Topology-aware MLPs (AAAI 2025 · similarity 0.87)
- Graph Disentangled Contrastive Learning with Personalized Transfer for Cross-Domain Recommendation (AAAI 2024 · 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.
