Research map / Label-efficient and robust learning / Recognition and long-tail
DMGIN: How Multimodal LLMs Enhance Large Recommendation Models for Lifelong User Post-click Behaviors
Zhuoxing Wei, Qingchen Xie, Qi Liu, Jingsong Yu
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 interest statistics features (ablation) | DMGIN-simple · Industry | AUC: 0.7129 -> 0.7134 | Improved |
| add intra-group behavior evolution (ablation) | DMGIN-simple · Industry | AUC: 0.7110 -> 0.7134 | Improved |
| add interest statistics features (ablation) | DMGIN-simple · Amazon | AUC: 0.7679 -> 0.7701 | Improved |
| add intra-group behavior evolution (ablation) | DMGIN-simple · Amazon | AUC: 0.7685 -> 0.7701 | Improved |
| add interest statistics features (ablation) | DMGIN-simple · Industry | GAUC: 0.6435 -> 0.6440 | Improved |
| add intra-group behavior evolution (ablation) | DMGIN-simple · Amazon | GAUC: 0.6984 -> 0.6986 | Improved |
| add temporal group evolution transformer (HSTU) on top of DMGIN-simple (ablation) | DMGIN · Industry | AUC: 0.7134 -> 0.7161 | Improved |
| add temporal group evolution transformer (HSTU) on top of DMGIN-simple (ablation) | DMGIN · Amazon | AUC: 0.7701 -> 0.7747 | Improved |
| increase TGETM depth from 1 to 6 HSTU layers (ablation) | DMGIN · Amazon | AUC: about 0.3%-0.5% AUC gain | Improved |
| use DMGIN instead of DSIN Full | DSIN Full · Industry | AUC: 0.7121 -> 0.7161 | Improved |
| use DMGIN instead of DSIN Full | DSIN Full · Amazon | AUC: 0.7663 -> 0.7747 | Improved |
| use full lifelong sequences (DIN Full) instead of short-sequence DIN | DIN · Amazon | AUC: 0.7245 -> 0.7634 | Improved |
Closest papers by meaning
- Harnessing Multimodal Large Language Models for Multimodal Sequential Recommendation (AAAI 2025 · similarity 0.88)
- LGMRec: Local and Global Graph Learning for Multimodal Recommendation (AAAI 2024 · similarity 0.88)
- Long-Sequence Recommendation Models Need Decoupled Embeddings (ICLR 2025 · similarity 0.87)
- Curriculum Conditioned Diffusion for Multimodal Recommendation (AAAI 2025 · similarity 0.87)
- LLM-ESR: Large Language Models Enhancement for Long-tailed Sequential Recommendation (NeurIPS 2024 · similarity 0.87)
- DOGE: LLMs-Enhanced Hyper-Knowledge Graph Recommender for Multimodal Recommendation (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.
