Adversarial Counterfactual Learning and Evaluation for Recommender System

TL;DR

This paper introduces an adversarial counterfactual learning framework for recommender systems that accounts for exposure bias, providing theoretical insights, a minimax formulation, and extensive simulations to improve recommendation accuracy.

Contribution

It proposes a novel minimax empirical risk formulation and an adversarial training approach to address exposure bias in recommender systems, with theoretical guarantees.

Findings

The approach improves recommendation accuracy under exposure bias.

Theoretical bounds support the effectiveness of the method.

Simulation results demonstrate benefits across various settings.

Abstract

The feedback data of recommender systems are often subject to what was exposed to the users; however, most learning and evaluation methods do not account for the underlying exposure mechanism. We first show in theory that applying supervised learning to detect user preferences may end up with inconsistent results in the absence of exposure information. The counterfactual propensity-weighting approach from causal inference can account for the exposure mechanism; nevertheless, the partial-observation nature of the feedback data can cause identifiability issues. We propose a principled solution by introducing a minimax empirical risk formulation. We show that the relaxation of the dual problem can be converted to an adversarial game between two recommendation models, where the opponent of the candidate model characterizes the underlying exposure mechanism. We provide learning bounds and conduct extensive simulation studies to illustrate and justify the proposed approach over a broad range of recommendation settings, which shed insights on the various benefits of the proposed approach.