Research map / Graph learning / Knowledge and temporal graphs
EdGCL: Disentangling Social and Cognitive Homophily in Graph-Based Educational Recommender Systems
Qingqing Liang, Chunyang Wang, Peiwei Xia, Yanan Zeng, Xin Liu, Xuesong Lu, Aoying Zhou
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-view contrastive disentanglement loss (ablation) | EdGCL · MOOPer | Recall@10: 0.6023 -> 0.6136 | Improved |
| add social link prediction auxiliary task (ablation) | EdGCL · MOOPer | Recall@10: 0.5936 -> 0.6136 | Improved |
| add cognitive diagnosis auxiliary task (ablation) | EdGCL · MOOPer | Recall@10: 0.5809 -> 0.6136 | Improved |
| add cross-view contrastive loss (ablation) | EdGCL · MOOCCubeX | Recall@10: 0.5507 -> 0.5576 | Improved |
| add social link prediction (ablation) | EdGCL · MOOCCubeX | Recall@10: 0.5391 -> 0.5576 | Improved |
| add cognitive diagnosis (ablation) | EdGCL · MOOCCubeX | Recall@10: 0.5489 -> 0.5576 | Improved |
| disentangle social and cognitive views with individual and community levels, versus Base, Cog., Cog.+Hom., Cog.+Soc. variants (ablation) | EdGCL · MOOCCubeX | NDCG@10: 0.5305 -> 0.5867 (Base vs EdGCL) | Improved |
| use graph transformer multi-hop aggregation with K=4 (MOOCCubeX) or 5 (MOOPer) instead of K limited to 2 or 3 (ablation) | EdGCL · MOOCCubeX and MOOPer hop length | Recall@10, NDCG@10 | Improved |
| model social homophily with EdGCL compared to graph and educational recommenders without social behaviors | EdGCL · MOOCCubeX | Recall@10: 0.5391 -> 0.5576 (HDSR best baseline) | Improved |
| use an LLM-based recommender (TALLRec) instead of graph models | Recommendation baselines · MOOCCubeX | NDCG@10: 0.2444 vs 0.5867 (EdGCL) | Hurt |
Closest papers by meaning
- Collaborative Cognitive Diagnosis with Disentangled Representation Learning for Learner Modeling (NeurIPS 2024 · similarity 0.87)
- Robust Graph Based Social Recommendation Through Contrastive Multi-View Learning (AAAI 2025 · similarity 0.85)
- Disentangled Modeling of Preferences and Social Influence for Group Recommendation (AAAI 2025 · similarity 0.85)
- Exploiting the Social-Like Prior in Transformer for Visual Reasoning (AAAI 2024 · similarity 0.84)
- iLLuMinaTE: An LLM-XAI Framework Leveraging Social Science Explanation Theories Towards Actionable Student Performance Feedback (AAAI 2025 · similarity 0.83)
- PLD: A Choice-Theoretic List-Wise Knowledge Distillation (NeurIPS 2025 · similarity 0.83)
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.
