Research map / Learning theory and algorithms / Prediction and decision losses
Removing Hidden Confounding in Recommendation: A Unified Multi-Task Learning Approach
Haoxuan Li, Kunhan Wu, Chunyuan Zheng, Yanghao Xiao, Hao Wang, Zhi Geng, Fuli Feng, Xiangnan He, Peng Wu
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 propensity and imputation residual networks to ESCM2-DR (Res-DR) (ablation) | ESCM2-DR · COAT | AUC: 0.760 -> 0.793 | Improved |
| add propensity and imputation residual networks to ESCM2-DR (Res-DR) (ablation) | ESCM2-DR · YAHOO! R3 | NDCG@5: 0.566 -> 0.654 | Improved |
| add residual networks and consistency loss to ESCM2-IPS (Res-IPS) (ablation) | ESCM2-IPS · KUAIREC | AUC: 0.815 -> 0.849 | Improved |
| remove both residual networks from Res-DR (ablation) | Res-DR · COAT | AUC: 0.793 -> 0.763 | Hurt |
| remove propensity residual network from Res-DR (ablation) | Res-DR · YAHOO! R3 | AUC: 0.750 -> 0.734 | Hurt |
| remove imputation residual network from Res-DR (ablation) | Res-DR · YAHOO! R3 | AUC: 0.750 -> 0.721 | Hurt |
| remove consistency loss L_B&U_CVR from Res-IPS (keep L_U_CVR) (ablation) | Res-IPS · COAT | AUC: 0.777 -> 0.770 | Hurt |
| remove L_U_CVR from Res-DR (keep L_B&U_CVR) (ablation) | Res-DR · COAT | AUC: 0.793 -> 0.790 | Hurt |
| remove L_B&U_CVR from Res-DR (keep L_U_CVR) (ablation) | Res-DR · YAHOO! R3 | NDCG@5: 0.654 -> 0.604 | Hurt |
| remove L_U_CVR from Res-IPS (ablation) | Res-IPS · COAT | AUC: 0.777 -> 0.784 | Mixed |
| increase unbiased data ratio from 2% to 10% | Res-DR · KUAIREC | AUC: 0.850 -> 0.855 | Improved |
| increase unbiased data ratio from 2% to 10% | CausE · KUAIREC | AUC: 0.818 -> 0.819 | No clear effect |
Closest papers by meaning
- Addressing Hidden Confounding with Heterogeneous Observational Datasets for Recommendation (NeurIPS 2024 · similarity 0.92)
- Alleviating Dual Biases in Recommendation (Student Abstract) (AAAI 2025 · similarity 0.88)
- Relaxing the Accurate Imputation Assumption in Doubly Robust Learning for Debiased Collaborative Filtering (ICML 2024 · similarity 0.87)
- CURE4Rec: A Benchmark for Recommendation Unlearning with Deeper Influence (NeurIPS 2024 · similarity 0.86)
- TranSUN: A Preemptive Paradigm to Eradicate Retransformation Bias Intrinsically from Regression Models in Recommender Systems (NeurIPS 2025 · similarity 0.86)
- Tree of Preferences for Diversified Recommendation (NeurIPS 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.
