Research map / Probabilistic and scientific modeling / Molecules and proteins
CORAL: Uncertainty-Aware Regulation of Exposure Concentration in Recommender Systems
Nitin Bisht, Linjiang Guo, Xiuwen Gong, Huan Huo, Guandong Xu
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 CORAL risk-aware regulation on top of SASRec | SASRec · Amazon overall EC | EC: 0.58 -> 0.20 | Improved |
| add CORAL regulation on top of Bert4Rec | Bert4Rec · Amazon overall EC | EC: 0.46 -> 0.18 | Improved |
| add CORAL regulation on top of SASRec | SASRec · Amazon utility | Recall@10: 0.1731 -> 0.1667 | Hurt |
| add CORAL regulation on top of SASRec | SASRec · ML-1M utility | Recall@10: 0.1916 -> 0.1768 | Hurt |
| use CORAL instead of Filter-style hard filtering | SASRec · Amazon utility | Recall@10: 0.0837 -> 0.1667 | Improved |
| add CORAL in the closed-loop environment with coarse categories | Naive recommender · Amazon closed-loop | SCC: 0.236 -> 0.344 | Improved |
| make the category taxonomy fine-grained (112 vs 28 categories) | CORAL · Amazon closed-loop | SCC: 0.344 -> 0.292 | Hurt |
| add CORAL when a mid-frequency category is held out | Naive recommender · ML-1M held-out category | CR: 0.328 -> 0.194 (t=1-33); 0.242 -> 0.283 (t=67-100) | Mixed |
| use Hawkes intensity as predictor of user rejection | Hawkes intensity model · Amazon rejection logistic regression | intensity coefficient beta1: 7.15 | Improved |
Closest papers by meaning
- Mitigating Confounding Bias in Practical Recommender Systems With Partially Inaccessible Exposure Status (TPAMI 2023 · similarity 0.85)
- CARE: Adaptive Calibration for Reliable Recommendations (ICML 2026 · similarity 0.85)
- Evidential Stochastic Differential Equations for Time-Aware Sequential Recommendation (NeurIPS 2024 · similarity 0.84)
- LLM4RSR: Large Language Models as Data Correctors for Robust Sequential Recommendation (AAAI 2025 · similarity 0.83)
- Influential Exemplar Replay for Incremental Learning in Recommender Systems (AAAI 2024 · similarity 0.83)
- On Efficiency-Effectiveness Trade-off of Diffusion-based Recommenders (NeurIPS 2025 · similarity 0.82)
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.
