paper-reading

Recommender Systems

Popularity Bias

Model

• PDA: Causal Intervention for Leveraging Popularity Bias in Recommendation. Xiangnan He et al. SIGIR, 2021. (PDF) (Source) (Code) (Slides) (Citations 217)

• TIDE: Popularity Bias Is Not Always Evil: Disentangling Benign and Harmful Bias for Recommendation. Xiangfan He et al. TKDE, 2023. (PDF) (Source) (Code) (Citations 30)

• MF-IPS: Recommendations as Treatments: Debiasing Learning and Evaluation. Schnabel T, Swaminathan A, Singh A, et al. ICML, 2016. (PDF)(Source)(Code)(Citation 638)

• DICE: Disentangling user interest and conformity for recommendation with causal embedding. Zheng Y, Gao C, Li X, et al. WWW, 2021. (PDF)(Source)(Code)(Citation 227)

• MACR: Model-agnostic counterfactual reasoning for eliminating popularity bias in recommender system. Wei T, Feng F, Chen J, et al. KDD, 2021. (PDF)(Source)(Code)(Citation 202)

• CauseE: Causal embeddings for recommendation. Bonner S, Vasile F. RecSys, 2018. (PDF)(Source)(Code)(Citation 245)

• DCCL：Disentangled Causal Embedding With Contrastive Learning For Recommender System（快手 2023）

• DCRS：Disentangled Representation for Diversified Recommendations（字节 2023）

• Self-supervised Learning for Large-scale Item Recommendations（Google 2020）

• ESAM：discriminative domain adaptation with non-displayed items to improve long-tail performance（Alibaba 2020）

• Co-training Disentangled Domain Adaptation Network for Leveraging Popularity Bias in Recommenders（Alibaba 2022）

Metrics

Surver

• On the Generalizability and Predictability of Recommender Systems. NeurIPS, 2022.
• Our Model Achieves Excellent Performance on MovieLens: What Does it Mean?
• Take a Fresh Look at Recommender Systems from an Evaluation Standpoint
• [ ]

Benchmark

Framework

Three-Way Decision

• TAO: Three-way decision and granular computing. Yao Y. IJAR, 2018. (PDF)(Source)(Code)(Citation 461)