Leveraging Geometric Insights in Hyperbolic Triplet Loss for Improved Recommendations

TL;DR

This paper introduces a hyperbolic recommendation model that leverages geometric insights to enhance representation learning, improve computational stability, and produce more diverse, personalized recommendations by modeling complex user-item interactions.

Contribution

It presents a novel hyperbolic triplet loss formulation that captures ternary relations and extends hyperbolic distance measures for better recommendation performance.

Findings

Outperforms existing Euclidean and hyperbolic models.

Reduces popularity bias in recommendations.

Enhances diversity and personalization of recommendations.

Abstract

Recent studies have demonstrated the potential of hyperbolic geometry for capturing complex patterns from interaction data in recommender systems. In this work, we introduce a novel hyperbolic recommendation model that uses geometrical insights to improve representation learning and increase computational stability at the same time. We reformulate the notion of hyperbolic distances to unlock additional representation capacity over conventional Euclidean space and learn more expressive user and item representations. To better capture user-items interactions, we construct a triplet loss that models ternary relations between users and their corresponding preferred and nonpreferred choices through a mix of pairwise interaction terms driven by the geometry of data. Our hyperbolic approach not only outperforms existing Euclidean and hyperbolic models but also reduces popularity bias, leading to more diverse and personalized recommendations.