Aligning Explanations for Recommendation with Rating and Feature via Maximizing Mutual Information

TL;DR

This paper introduces a mutual information-based framework to improve the alignment of natural language explanations with predicted ratings and item features in recommendation systems, enhancing user trust and decision-making.

Contribution

The proposed MMI framework is a flexible, model-agnostic method that uses mutual information to better align explanations with ratings and features, outperforming existing methods.

Findings

Boosts alignment with predicted ratings and features

Outperforms baseline models in experiments

User studies confirm improved decision support

Abstract

Providing natural language-based explanations to justify recommendations helps to improve users' satisfaction and gain users' trust. However, as current explanation generation methods are commonly trained with an objective to mimic existing user reviews, the generated explanations are often not aligned with the predicted ratings or some important features of the recommended items, and thus, are suboptimal in helping users make informed decision on the recommendation platform. To tackle this problem, we propose a flexible model-agnostic method named MMI (Maximizing Mutual Information) framework to enhance the alignment between the generated natural language explanations and the predicted rating/important item features. Specifically, we propose to use mutual information (MI) as a measure for the alignment and train a neural MI estimator. Then, we treat a well-trained explanation generation model as the backbone model and further fine-tune it through reinforcement learning with guidance from the MI estimator, which rewards a generated explanation that is more aligned with the predicted rating or a pre-defined feature of the recommended item. Experiments on three datasets demonstrate that our MMI framework can boost different backbone models, enabling them to outperform existing baselines in terms of alignment with predicted ratings and item features. Additionally, user studies verify that MI-enhanced explanations indeed facilitate users' decisions and are favorable compared with other baselines due to their better alignment properties.