# Popularity Bias Is Not Always Evil: Disentangling Benign and Harmful Bias for Recommendation

arXiv: https://arxiv.org/abs/2109.07946 · HTML: https://mlsanity.com/paper/2109.07946/

## Key facts

- Cited by 23 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 16 Sep 2021.

## Built on by (most cited first, 23 of 23)

- [A Survey on Popularity Bias in Recommender Systems](https://mlsanity.com/paper/2308.01118/) (2023)
- [Fairness in Recommender Systems: Research Landscape and Future Directions](https://mlsanity.com/paper/2205.11127/) (2022)
- [Large Language Models as Recommender Systems: A Study of Popularity Bias](https://mlsanity.com/paper/2406.01285/) (2024)
- [How Do Recommendation Models Amplify Popularity Bias? An Analysis from the Spectral Perspective](https://mlsanity.com/paper/2404.12008/) (2024)
- [Improving Conversational Recommendation Systems via Bias Analysis and Language-Model-Enhanced Data Augmentation](https://mlsanity.com/paper/2310.16738/) (2023)
- [PSL: Rethinking and Improving Softmax Loss from Pairwise Perspective for Recommendation](https://mlsanity.com/paper/2411.00163/) (2024)
- [Heterophily-Aware Fair Recommendation using Graph Convolutional Networks](https://mlsanity.com/paper/2402.03365/) (2024)
- [Beyond Single Labels: Improving Conversational Recommendation through LLM-Powered Data Augmentation](https://mlsanity.com/paper/2508.05657/) (2025)
- [Metrics for popularity bias in dynamic recommender systems](https://mlsanity.com/paper/2310.08455/) (2023)
- [Uncertainty Calibration for Counterfactual Propensity Estimation in Recommendation](https://mlsanity.com/paper/2303.12973/) (2023)
- [CDR: Conservative Doubly Robust Learning for Debiased Recommendation](https://mlsanity.com/paper/2308.08461/) (2023)
- [Algorithmic Fairness: A Tolerance Perspective](https://mlsanity.com/paper/2405.09543/) (2024)
- [Simulating News Recommendation Ecosystem for Fun and Profit](https://mlsanity.com/paper/2305.14103/) (2023)
- [Capturing Popularity Trends: A Simplistic Non-Personalized Approach for Enhanced Item Recommendation](https://mlsanity.com/paper/2308.08799/) (2023)
- [Debiased Contrastive Representation Learning for Mitigating Dual Biases in Recommender Systems](https://mlsanity.com/paper/2408.09646/) (2024)
- [The Importance of Causality in Decision Making: A Perspective on Recommender Systems](https://mlsanity.com/paper/2410.01822/) (2024)
- [Fairness Aware Reward Optimization](https://mlsanity.com/paper/2602.07799/) (2026)
- [Hidden Author Bias in Book Recommendation](https://mlsanity.com/paper/2209.00371/) (2022)
- [Ranking with Popularity Bias: User Welfare under Self-Amplification Dynamics](https://mlsanity.com/paper/2305.18333/) (2023)
- [Teleporter Theory: A General and Simple Approach for Modeling Cross-World Counterfactual Causality](https://mlsanity.com/paper/2406.11501/) (2024)
- [Taming Recommendation Bias with Causal Intervention on Evolving Personal Popularity](https://mlsanity.com/paper/2505.14310/) (2025)
- [BPL: Bias-adaptive Preference Distillation Learning for Recommender System](https://mlsanity.com/paper/2510.16076/) (2025)
- [Debiasing Message Passing to Mitigate Popularity Bias in GNN-based Collaborative Filtering](https://mlsanity.com/paper/2605.11145/) (2026)

## Abstract (arXiv)

Recommender system usually suffers from severe popularity bias -- the collected interaction data usually exhibits quite imbalanced or even long-tailed distribution over items. Such skewed distribution may result from the users' conformity to the group, which deviates from reflecting users' true preference. Existing efforts for tackling this issue mainly focus on completely eliminating popularity bias. However, we argue that not all popularity bias is evil. Popularity bias not only results from conformity but also item quality, which is usually ignored by existing methods. Some items exhibit higher popularity as they have intrinsic better property. Blindly removing the popularity bias would lose such important signal, and further deteriorate model performance. To sufficiently exploit such important information for recommendation, it is essential to disentangle the benign popularity bias caused by item quality from the harmful popularity bias caused by conformity. Although important, it is quite challenging as we lack an explicit signal to differentiate the two factors of popularity bias. In this paper, we propose to leverage temporal information as the two factors exhibit quite different patterns along the time: item quality revealing item inherent property is stable and static while conformity that depends on items' recent clicks is highly time-sensitive. Correspondingly, we further propose a novel Time-aware DisEntangled framework (TIDE), where a click is generated from three components namely the static item quality, the dynamic conformity effect, as well as the user-item matching score returned by any recommendation model. Lastly, we conduct interventional inference such that the recommendation can benefit from the benign popularity bias while circumvent the harmful one. Extensive experiments on three real-world datasets demonstrated the effectiveness of TIDE.