Accelerated Convergence for Counterfactual Learning to Rank

11Citations
Citations of this article
30Readers
Mendeley users who have this article in their library.
Get full text

Abstract

Counterfactual Learning To Rank (LTR) algorithms learn a ranking model from logged user interactions, often collected using a production system. Employing such an offline learning approach has many benefits compared to an online one, but it is challenging as user feedback often contains high levels of bias. Unbiased LTR uses Inverse Propensity Scoring (IPS) to enable unbiased learning from logged user interactions. One of the major difficulties in applying Stochastic Gradient Descent (SGD) approaches to counterfactual learning problems is the large variance introduced by the propensity weights. In this paper we show that the convergence rate of SGD approaches with IPS-weighted gradients suffers from the large variance introduced by the IPS weights: convergence is slow, especially when there are large IPS weights. To overcome this limitation, we propose a novel learning algorithm, called CounterSample, that has provably better convergence than standard IPS-weighted gradient descent methods. We prove that CounterSample converges faster and complement our theoretical findings with empirical results by performing extensive experimentation in a number of biased LTR scenarios-across optimizers, batch sizes, and different degrees of position bias.

References Powered by Scopus

Cumulated gain-based evaluation of IR techniques

3598Citations
N/AReaders
Get full text

Understanding machine learning: From theory to algorithms

3309Citations
N/AReaders
Get full text

Optimizing search engines using clickthrough data

3074Citations
N/AReaders
Get full text

Cited by Powered by Scopus

Adapting Interactional Observation Embedding for Counterfactual Learning to Rank

17Citations
N/AReaders
Get full text

Overview of the Frontier Progress of Causal Machine Learning

9Citations
N/AReaders
Get full text

Rax: Composable Learning-to-Rank Using JAX

7Citations
N/AReaders
Get full text

Register to see more suggestions

Mendeley helps you to discover research relevant for your work.

Already have an account?

Cite

CITATION STYLE

APA

Jagerman, R., & De Rijke, M. (2020). Accelerated Convergence for Counterfactual Learning to Rank. In SIGIR 2020 - Proceedings of the 43rd International ACM SIGIR Conference on Research and Development in Information Retrieval (pp. 469–478). Association for Computing Machinery, Inc. https://doi.org/10.1145/3397271.3401069

Readers' Seniority

Tooltip

PhD / Post grad / Masters / Doc 10

71%

Researcher 3

21%

Professor / Associate Prof. 1

7%

Readers' Discipline

Tooltip

Computer Science 13

81%

Business, Management and Accounting 1

6%

Environmental Science 1

6%

Mathematics 1

6%

Save time finding and organizing research with Mendeley

Sign up for free