LLMs for User Interest Exploration in Large-scale Recommendation Systems

TL;DR

This paper proposes a hybrid hierarchical framework combining LLMs and traditional recommendation models to enhance user interest exploration and discover novel interests in large-scale recommendation systems.

Contribution

It introduces a novel hierarchical framework with interest clusters and language-based interest descriptions to improve exploration beyond feedback loops in recommendation systems.

Findings

Significant increase in novel interest exploration

Enhanced user enjoyment on a commercial platform

Effective control over interest granularity

Abstract

Traditional recommendation systems are subject to a strong feedback loop by learning from and reinforcing past user-item interactions, which in turn limits the discovery of novel user interests. To address this, we introduce a hybrid hierarchical framework combining Large Language Models (LLMs) and classic recommendation models for user interest exploration. The framework controls the interfacing between the LLMs and the classic recommendation models through "interest clusters", the granularity of which can be explicitly determined by algorithm designers. It recommends the next novel interests by first representing "interest clusters" using language, and employs a fine-tuned LLM to generate novel interest descriptions that are strictly within these predefined clusters. At the low level, it grounds these generated interests to an item-level policy by restricting classic recommendation models, in this case a transformer-based sequence recommender to return items that fall within the novel clusters generated at the high level. We showcase the efficacy of this approach on an industrial-scale commercial platform serving billions of users. Live experiments show a significant increase in both exploration of novel interests and overall user enjoyment of the platform.