A Complete Framework for Offline and Counterfactual Evaluations of Interactive Recommendation Systems
Open Access
- 23 October 2023
- conference paper
- Published by Association for Computing Machinery (ACM)
- p. 193-197
Abstract
Interactive recommendation has been recognized as a Multi-Armed Bandit (MAB) problem. Items are arms to be pulled (i.e., recommended) and the user’s satisfaction is the reward to be maximized. Despite the advances, there is still a lack of consensus on the best practices to evaluate such solutions. Recently, two complementary frameworks were proposed to evaluate bandit solutions more accurately: iRec and OBP. The first one has a complete set of offline metrics and bandit models that allows us to perform an comparisons with several evaluation policies. The second one provides a huge set of bandit models to be evaluated through several counterfactual estimators. However, there is a room to be explored when joining these two frameworks. We propose and evaluate an integration between both, demonstrating the potential and richness of such combination.Keywords
This publication has 18 references indexed in Scilit:
- Neural Interactive Collaborative Filtering2020
- Interactive Recommender System via Knowledge Graph-enhanced Reinforcement Learning2020
- A simple multi-armed nearest-neighbor bandit for interactive recommendation2019
- Online Interactive Collaborative Filtering Using Multi-Armed Bandit with Dependent ArmsIEEE Transactions on Knowledge and Data Engineering, 2018
- Learning Contextual Bandits in a Non-stationary Environment2018
- Linear Thompson sampling revisitedElectronic Journal of Statistics, 2017
- Collaborative Filtering Bandits2016
- Interactive collaborative filtering2013
- Finite-time Analysis of the Multiarmed Bandit ProblemMachine Learning, 2002
- 10.1162/153244303321897663Applied Physics Letters, 2000