Adversarial Counterfactual Learning and Evaluation for Recommender System
This addresses bias in recommender systems for users and platforms, offering a principled method to improve accuracy, but it is incremental as it builds on existing causal inference techniques.
The paper tackles the problem of learning and evaluating recommender systems when feedback data is biased by exposure mechanisms, showing that standard supervised learning can lead to inconsistent results. It proposes a counterfactual propensity-weighting solution using an adversarial game between models, with theoretical bounds and simulation studies demonstrating benefits across various settings.
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.