Research map / Label-efficient and robust learning / Recognition and long-tail
Theoretically Guaranteed Bidirectional Data Rectification for Robust Sequential Recommendation
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 DRUT (target rectification) to SASRec (ablation) | SASRec · four datasets (avg) | NDCG@10: 45.7% | Helped |
| add DDUI (input rectification) to SASRec (ablation) | SASRec · four datasets (avg) | NDCG@10: 43.3% | Helped |
| add self-ensemble on top of DRUT + DDUI (ablation) | BirDRec · four datasets (avg) | NDCG@10: 3.1% | Helped |
| use DRUT and DDUI together instead of either alone (ablation) | SASRec · four datasets | NDCG@10 | Helped |
| apply BirDRec to FPMC backbone | FPMC · ML-1M | NDCG@10: 0.1069 -> 0.1251 | Helped |
| apply BirDRec to GRU4Rec backbone | GRU4Rec · Yelp | NDCG@10: 0.0265 -> 0.0659 | Helped |
| apply BirDRec to MAGNN backbone | MAGNN · Beauty | NDCG@10: 0.0455 -> 0.0518 | Helped |
| BirDRec (SASRec) vs best robust baseline BERD | BirDRec · ML-1M | NDCG@10: 0.1554 -> 0.1915 | Helped |
| BirDRec vs STEAM on training speed | BirDRec · time per epoch | speedup: 4.28 times faster | Helped |
| BirDRec memory vs STEAM on ML-1M (small item set) | BirDRec · ML-1M | GPU memory | Mixed |
| BirDRec training cost vs backbone | BirDRec · per-epoch training time | time: 1.6 times slower | Hurt |
| rectification thresholds beta and beta' set to 0.1 rather than larger (ablation) | BirDRec · ML-1M | NDCG@10 | Helped |
Closest papers by meaning
- LLM4RSR: Large Language Models as Data Correctors for Robust Sequential Recommendation (AAAI 2025 · similarity 0.89)
- Unbiased Rectification for Sequential Recommender Systems Under Fake Orders (AAAI 2026 · similarity 0.88)
- Augmenting Sequential Recommendation with Balanced Relevance and Diversity (AAAI 2025 · similarity 0.87)
- Future Sight and Tough Fights: Revolutionizing Sequential Recommendation with FENRec (AAAI 2025 · similarity 0.87)
- On Efficiency-Effectiveness Trade-off of Diffusion-based Recommenders (NeurIPS 2025 · similarity 0.85)
- SLMRec: Distilling Large Language Models into Small for Sequential Recommendation (ICLR 2025 · similarity 0.85)
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.