paper-with-me

홈 › Papers

Unbiased Learning to Rank Meets Reality: Lessons from Baidu's Large-Scale Search Dataset

2024-04-03 · Philipp Hager, Romain Deffayet, Jean-Michel Renders, Onno Zoeter, Maarten de Rijke

Unbiased learning-to-rank (ULTR) is a well-established framework for learning from user clicks, which are often biased by the ranker collecting the data. While theoretically justified and extensively tested in simulation, ULTR techniques lack empirical validation, especially on modern search engines. The Baidu-ULTR dataset released for the WSDM Cup 2023, collected from Baidu's search engine, offers a rare opportunity to assess the real-world performance of prominent ULTR techniques. Despite multiple submissions during the WSDM Cup 2023 and the subsequent NTCIR ULTRE-2 task, it remains unclear whether the observed improvements stem from applying ULTR or other learning techniques. In this work, we revisit and extend the available experiments on the Baidu-ULTR dataset. We find that standard unbiased learning-to-rank techniques robustly improve click predictions but struggle to consistently improve ranking performance, especially considering the stark differences obtained by choice of ranking loss and query-document features. Our experiments reveal that gains in click prediction do not necessarily translate to enhanced ranking performance on expert relevance annotations, implying that conclusions strongly depend on how success is measured in this benchmark.

📄 PDF Abstract BibTeX arXiv:2404.02543

Code (2)

philipphager/baidu-bert-model 공식 구현 jax
philipphager/ultr-reproducibility 공식 구현 jax

Tasks

Learning-To-Rank

Similar Papers 제목 키워드 기반

A Large Scale Search Dataset for Unbiased Learning to Rank

2022-07-07 · Lixin Zou, Haitao Mao, Xiaokai Chu, Jiliang Tang 외

The unbiased learning to rank (ULTR) problem has been greatly advanced by recent deep learning techniques and well-designed debias algorithms. However, promising results on the existing benchmark datasets may not be exte…

Causal DiscoveryLanguage ModellingLearning-To-RankMeta-Learning+1

Understanding the Effects of the Baidu-ULTR Logging Policy on Two-Tower Models

2024-09-18 · Morris de Haan, Philipp Hager

Despite the popularity of the two-tower model for unbiased learning to rank (ULTR) tasks, recent work suggests that it suffers from a major limitation that could lead to its collapse in industry applications: the problem…

Learning-To-Rank

Contextual Dual Learning Algorithm with Listwise Distillation for Unbiased Learning to Rank

2024-08-19 · Lulu Yu, Keping Bi, Shiyu Ni, Jiafeng Guo

Unbiased Learning to Rank (ULTR) aims to leverage biased implicit user feedback (e.g., click) to optimize an unbiased ranking model. The effectiveness of the existing ULTR methods has primarily been validated on syntheti…

Learning-To-RankPosition

Unbiased Learning to Rank with Query-Level Click Propensity Estimation: Beyond Pointwise Observation and Relevance

2025-02-17 · Lulu Yu, Keping Bi, Jiafeng Guo, Shihao Liu 외

Most existing unbiased learning-to-rank (ULTR) approaches are based on the user examination hypothesis, which assumes that users will click a result only if it is both relevant and observed (typically modeled by position…

Learning-To-RankPosition

Multi-Feature Integration for Perception-Dependent Examination-Bias Estimation

2023-02-27 · Xiaoshu Chen, Xiangsheng Li, Kunliang Wei, Bin Hu 외

Eliminating examination bias accurately is pivotal to apply click-through data to train an unbiased ranking model. However, most examination-bias estimators are limited to the hypothesis of Position-Based Model (PBM), wh…