Research map / Graph learning / Graph self-supervised learning
Subspace-Aware Graph Construction and Contrastive Alignment for Multimodal Recommendation with Large Language Models
Haodong Li, Lianyong Qi, Weiming Liu, Fan Wang, Chong Li, Shengye Pang, Wenwen Gong, Yanwei Xu, Xiaoxiao Chi, Yang Zhang, Xiaokang 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 LLM-generated profiles (ablation) | SCALE · Amazon Baby | Recall@20 | Improved |
| add semantic alignment loss (ablation) | SCALE · Amazon Baby | Recall@20 | Improved |
| add relation alignment loss (ablation) | SCALE · Amazon Baby | Recall@20 | Improved |
| add subspace-aware graph (OMP) (ablation) | SCALE · Amazon Baby | Recall@20 | Improved |
| add feature similarity graph (ablation) | SCALE · Amazon Baby | Recall@20 | Improved |
| add inter-entity user-item semantic relations (ablation) | SCALE · Amazon Baby | Recall@20 | Improved |
| add intra-entity user-user and item-item relations (ablation) | SCALE · Amazon Baby | Recall@20 | Improved |
| increase semantic graph convolution layers (ablation) | SCALE · Amazon Baby | Recall@20 | Hurt |
| increase number of negative samples in relation alignment loss (ablation) | SCALE · Baby, Book, Yelp | Recall@20 | No clear effect |
| use sparsity level 10 instead of too small or too large (ablation) | SCALE · Amazon Baby | Recall@20 | Mixed |
| use SCALE instead of RLMRec (SGL) on Book | RLMRec · Book | Recall@20: 0.1537 -> 0.1624 | Improved |
| use SCALE instead of MENTOR on Baby | MENTOR · Amazon Baby | Recall@20: 0.1048 -> 0.1341 | Improved |
Closest papers by meaning
- Beyond Graph Convolution: Multimodal Recommendation with Topology-aware MLPs (AAAI 2025 · similarity 0.90)
- LGMRec: Local and Global Graph Learning for Multimodal Recommendation (AAAI 2024 · similarity 0.90)
- DOGE: LLMs-Enhanced Hyper-Knowledge Graph Recommender for Multimodal Recommendation (AAAI 2025 · similarity 0.89)
- Structured Spectral Reasoning for Frequency-Adaptive Multimodal Recommendation (NeurIPS 2025 · similarity 0.89)
- Seeing Beyond Noise: Joint Graph Structure Evaluation and Denoising for Multimodal Recommendation (AAAI 2025 · similarity 0.89)
- Modality-Independent Graph Neural Networks with Global Transformers for Multimodal Recommendation (AAAI 2025 · similarity 0.88)
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.
