Research map / Graph learning / Knowledge and temporal graphs
MoToRec: Sparse-Regularized Multimodal Tokenization for Cold-Start Recommender
Jialin Liu, Zhaorui Zhang, Ray C.C. Cheung
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 sparse-regularized RQ-VAE tokenization (ablation) | MoToRec · Amazon Baby (overall) | Recall@20: 0.0915 -> 0.1077 | Improved |
| add sparse-regularized RQ-VAE tokenization (ablation) | MoToRec · Amazon Baby (cold-start) | Recall@20: 0.0229 -> 0.0347 | Improved |
| add adaptive rarity amplification (ablation) | MoToRec · Amazon Baby (cold-start) | Recall@20: 0.0281 -> 0.0347 | Improved |
| add sparsity regularization (ablation) | MoToRec · Amazon Baby (cold-start) | Recall@20: 0.0277 -> 0.0347 | Improved |
| add contrastive loss (ablation) | MoToRec · Amazon Baby (overall) | Recall@20: 0.1026 -> 0.1077 | Improved |
| add hybrid fusion (ablation) | MoToRec · Amazon Baby (overall) | Recall@20: 0.1028 -> 0.1077 | Improved |
| add homogeneous graph enhancement (ablation) | MoToRec · Amazon Baby (overall) | Recall@20: 0.1002 -> 0.1077 | Improved |
| MoToRec vs best multimodal baselines | MoToRec · Amazon Baby/Sports/Clothing | Recall@20: Baby R@20 0.0989 -> 0.1077 | Improved |
| discrete tokenization on cold-start items | MoToRec · Amazon cold-start item set | NDCG@20: up to 12.58% N@20 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- LGMRec — LGMRec: Local and Global Graph Learning for Multimodal Recommendation (AAAI 2024) · better on 6 of 6 shared results
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- Beyond Graph Convolution: Multimodal Recommendation with Topology-aware MLPs (AAAI 2025 · similarity 0.88)
- DOGE: LLMs-Enhanced Hyper-Knowledge Graph Recommender for Multimodal Recommendation (AAAI 2025 · similarity 0.87)
- LGMRec: Local and Global Graph Learning for Multimodal Recommendation (AAAI 2024 · similarity 0.86)
- Improving LLMs for Recommendation with Out-Of-Vocabulary Tokens (ICML 2025 · similarity 0.86)
- Temporally and Distributionally Robust Optimization for Cold-Start Recommendation (AAAI 2024 · similarity 0.86)
- Semantic Convergence: Harmonizing Recommender Systems via Two-Stage Alignment and Behavioral Semantic Tokenization (AAAI 2025 · similarity 0.86)
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.
