Counterfactual Explanations for Neural Recommenders

TL;DR

This paper introduces ACCENT, a novel framework that generates tangible counterfactual explanations for neural recommender systems by identifying influential training points and iteratively deducing counterfactuals, enhancing interpretability.

Contribution

The paper presents the first general method for producing counterfactual explanations for neural recommenders, extending influence functions to pairs of items and applying it to popular models.

Findings

Successfully applied to NCF and RCF models

Demonstrated on MovieLens 100K dataset

Improves interpretability of neural recommenders

Abstract

Understanding why specific items are recommended to users can significantly increase their trust and satisfaction in the system. While neural recommenders have become the state-of-the-art in recent years, the complexity of deep models still makes the generation of tangible explanations for end users a challenging problem. Existing methods are usually based on attention distributions over a variety of features, which are still questionable regarding their suitability as explanations, and rather unwieldy to grasp for an end user. Counterfactual explanations based on a small set of the user's own actions have been shown to be an acceptable solution to the tangibility problem. However, current work on such counterfactuals cannot be readily applied to neural models. In this work, we propose ACCENT, the first general framework for finding counterfactual explanations for neural recommenders. It extends recently-proposed influence functions for identifying training points most relevant to a recommendation, from a single to a pair of items, while deducing a counterfactual set in an iterative process. We use ACCENT to generate counterfactual explanations for two popular neural models, Neural Collaborative Filtering (NCF) and Relational Collaborative Filtering (RCF), and demonstrate its feasibility on a sample of the popular MovieLens 100K dataset.