Explainable graph neural networks identify key peers in recommendations
xLightGCN: Explainable Graph Neural Network Based Recommendation Systems with Most Influential Peers
Artificial Intelligence
Summary
Recommendation systems suggest items you might like, but many use complex models that are hard to understand. This paper introduces a way to clearly explain why recommendations are made by showing which other users most influenced the suggestions. The authors build on a popular graph-based model to provide explanations with strong mathematical guarantees. This helps build trust in recommendation systems by making them less like black boxes.
What this means in practice
- •For social media engineers: Create recommendation systems that show exactly which user connections influenced each suggested item, improving transparency and user trust.
- •For online marketplace developers: Build recommendation features that provide interpretable reasons for suggestions by identifying the most influential peers in the purchase network.
Authors
Calder Johnson, Robin Cohen
Abstract
Recommendation systems, systems that predict a user's preference for items in order to curate that user's feed, are a ubiquitous and central feature in online social networks. While they are a central piece of our modern digital infrastructure, many modern approaches are opaque black boxes, making them difficult to audit, understand, and assess the trustworthiness of. In this paper we present an exact, graph-based approach for generating explanations for decisions made by recommendation systems. We argue that in the risk-sensitive sociotechnical context of user-facing AI such as recommendation systems, explainability is a key pillar for establishing the trustworthiness of models and validating their effectiveness. We propose a technique built on LightGCN like architectures, an influential and popular class of recommendation models. We describe our approach and argue that it provides an appropriate assurance mechanism for these AI systems leading to specific advantages, including interpretable explanations and a theoretically guaranteed measure of explanation fidelity. We conclude by discussing limitations and directions for future research.