Pessimistic Off-Policy Optimization for Learning to Rank

Čief, Matej; Kompan, Michal

Pessimistic Off-Policy Optimization for Learning to Rank

Files

FAIA392FAIA240703.pdf(591.37 KB)

Date

2024-10-21

Authors

Čief, Matej

Kompan, Michal

ORCID

0000-0002-4649-5120

Publisher

IOS Press

Altmetrics

Abstract

Off-policy learning is a framework for optimizing policies without deploying them, using data collected by another policy. In recommender systems, this is especially challenging due to the imbalance in logged data: some items are recommended and thus logged more frequently than others. This is further perpetuated when recommending a list of items, as the action space is combinatorial. To address this challenge, we study pessimistic off-policy optimization for learning to rank. The key idea is to compute lower confidence bounds on parameters of click models and then return the list with the highest pessimistic estimate of its value. This approach is computationally efficient, and we analyze it. We study its Bayesian and frequentist variants and overcome the limitation of unknown prior by incorporating empirical Bayes. To show the empirical effectiveness of our approach, we compare it to off-policy optimizers that use inverse propensity scores or neglect uncertainty. Our approach outperforms all baselines and is both robust and general.

Keywords

Action spaces, Bayesian, Computationally efficient, Empirical Bayes, Frequentist, Lower confidence bound, Optimizers, Optimizing policies, Policy learning, Policy optimization

Citation

27TH EUROPEAN CONFERENCE ON ARTIFICIAL INTELLIGENCE. 2024, p. 1896-1903.
https://ebooks.iospress.nl/volumearticle/69798

Document type

Peer-reviewed

Document version

Published version

Language of document

en

Document licence

Creative Commons Attribution-NonCommercial 4.0 International
http://creativecommons.org/licenses/by-nc/4.0/