Paper Info
Title
Revisiting Two-tower Models for Unbiased Learning to Rank
Abstract
Two-tower architecture is commonly used in real-world systems for Unbiased Learning to Rank (ULTR), where a Deep Neural Network (DNN) tower models unbiased relevance predictions, while another tower models observation biases inherent in the training data like user clicks. This two-tower architecture introduces inductive biases to allow more efficient use of limited observational logs and better generalization during deployment than single-tower architecture that may learn spurious correlations between relevance predictions and biases. However, despite their popularity, it is largely neglected in the literature that existing two-tower models assume that the joint distribution of relevance prediction and observation probabilities are completely factorizable. In this work, we revisit two-tower models for ULTR. We rigorously show that the factorization assumption can be too strong for real-world user behaviors, and existing methods may easily fail under slightly milder assumptions. We then propose several novel ideas that consider a wider spectrum of user behaviors while still under the two-tower framework to maintain simplicity and generalizability. Our concerns of existing two-tower models and the effectiveness of our proposed methods are validated on both controlled synthetic and large-scale real-world datasets.
Year
DOI
Venue
2022
10.1145/3477495.3531837
SIGIR '22: Proceedings of the 45th International ACM SIGIR Conference on Research and Development in Information Retrieval
Keywords
DocType
Citations 
Unbiased Learning to Rank, Expectation Maximization, Bias Factorization
Conference
1
PageRank 
References 
Authors
0.35
19
6
Name
Order
Citations
PageRank
Le Yan110.35
Zhen Qin213816.93
Honglei Zhuang319316.37
Xuanhui Wang4139468.85
Michael Bendersky598648.69
Marc A. Najork62538278.16