跳到正文
arXiv cs.AI· Blake Gella, Wei Wu, Yuhao Yin, Zexi Huang, Zikai Wang, Emily Liu, Junlin Zhang, Wentao Guo, Qinglei Wang·· 2 小时前AI 评分40

PEARL:基于对比学习的无偏分位数估计,用于工业级直播推荐

PEARL: Unbiased Percentile Estimation via Contrastive Learning for Industrial-Scale Livestream Recommendation

AI 导读

PEARL 是一个非参数对比分位数近似框架,通过建模相对偏好信号而非绝对互动强度,纠正用户行为强度失衡带来的偏差,理论证明成对比较可得到分位数偏好信号的无偏估计。

正文 · 原文

View PDF HTML (experimental)

Abstract:Recommender systems trained on user interaction data are susceptible to behavioral intensity imbalance--a systematic distortion arising from heterogeneous engagement patterns across users. This imbalance skews feedback signals such that observed interactions no longer faithfully reflect true preferences, causing models to disproportionately amplify signals from highly active users while underrepresenting others, which ultimately degrades recommendation quality and robustness at scale. To address this issue, we propose a nonparametric contrastive percentile approximation framework, PEARL, that models relative preference signals instead of absolute engagement magnitudes. Building upon relative advantage debiasing, PEARL leverages real contrastive interaction samples to approximate percentile relationships directly, without relying on auxiliary distribution estimation models. We provide theoretical justification demonstrating that such pairwise comparisons yield unbiased estimates of percentile-based preference signals. For broader applicability, we introduce a prediction-based bootstrapping mechanism for percentile smoothing to handle sparse and discrete feedback, alongside a generalized value-weighted formulation and a co-training strategy to enhance both modeling flexibility and representation learning. Extensive offline experiments demonstrate that PEARL effectively mitigates behavioral bias and consistently improves recommendation performance across multiple ranking targets. Deployed in a production livestream platform with a combined user base of billions, online A/B testing confirms substantial real-world gains: +2.10% Watch Duration, +0.80% Consumption Amount, +1.49% Interaction Rate, and -6.91% Report Rate.
Comments: Accepted to KDD 2027 ADS Track
Subjects: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
Cite as: arXiv:2605.21752 [cs.LG]
  (or arXiv:2605.21752v2 [cs.LG] for this version)
  https://doi.org/10.48550/arXiv.2605.21752

arXiv-issued DOI via DataCite

Submission history

From: Blake Gella [view email]
[v1] Wed, 20 May 2026 21:25:41 UTC (132 KB)
[v2] Tue, 17 Nov 2026 03:24:35 UTC (132 KB)

来源:arXiv cs.AI · arxiv.org