Research map / Label-efficient and robust learning / Recognition and long-tail
Hard vs. Noise: Resolving Hard-Noisy Sample Confusion in Recommender Systems via Large Language Models
Tianrui Song, Wen-Shuo Chao, Hao Liu
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-embedding guided hard negative sampling (ablation) | LLMHNI (LightGCN) · Amazon-book | Recall@10: 0.1199 -> 0.1423 | Improved |
| add objective alignment of LLM-encoded embeddings (ablation) | LLMHNI (LightGCN) · Amazon-book | Recall@10: 0.1248 -> 0.1423 | Improved |
| add hallucination-robust graph contrastive loss (ablation) | LLMHNI (LightGCN) · Amazon-book | Recall@10: 0.1125 -> 0.1423 | Improved |
| add cross-graph denoising contrastive loss (ablation) | LLMHNI (LightGCN) · Amazon-book | Recall@10: 0.1294 -> 0.1423 | Improved |
| add hallucination-robust graph contrastive loss (ablation) | LLMHNI (LightGCN) · Amazon-book | NDCG@10: 0.0855 -> 0.1168 | Improved |
| add hard negative sampling guided by semantic relevance (ablation) | LLMHNI (LightGCN) · Amazon-book | NDCG@20: 0.1112 -> 0.1369 | Improved |
| use LLMHNI instead of LLaRD (LightGCN) | LightGCN · Amazon-book | Recall@10: 0.1408 -> 0.1423 | Improved |
| use LLMHNI instead of vanilla training | LightGCN · Amazon-book | Recall@10: 0.0950 -> 0.1423 | Improved |
| use LLMHNI instead of vanilla training | NGCF · Amazon-book | Recall@10: 0.0763 -> 0.1290 | Improved |
| use LLMHNI instead of LLaRD | LightGCN · Yelp | Recall@10: 0.0975 -> 0.0981 | Improved |
| add 5-20% non-existent training interactions | LLMHNI · Amazon-books and Yelp | Recall | Improved |
| vary tau_hal and tau_de over 0.1-1.0 | LLMHNI · Amazon-books | Recall and NDCG | Mixed |
Closest papers by meaning
- LLM-ESR: Large Language Models Enhancement for Long-tailed Sequential Recommendation (NeurIPS 2024 · similarity 0.88)
- Can LLMs Outshine Conventional Recommenders? A Comparative Evaluation (NeurIPS 2025 · similarity 0.87)
- Language Ranker: A Lightweight Ranking framework for LLM Decoding (NeurIPS 2025 · similarity 0.87)
- Seeing Beyond Noise: Joint Graph Structure Evaluation and Denoising for Multimodal Recommendation (AAAI 2025 · similarity 0.87)
- FACE: A General Framework for Mapping Collaborative Filtering Embeddings into LLM Tokens (NeurIPS 2025 · similarity 0.87)
- Language Representations Can be What Recommenders Need: Findings and Potentials (ICLR 2025 · 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.
