Graph Meets LLM: A Novel Approach to Collaborative Filtering for Robust Conversational Understanding

Abstract

A Personalized Query Rewriting system aims to reduce defective queries to ensure robust conversational functionality by considering individual user behavior and preferences. It's usually structured as a search-based system, maintaining a user history index of past successful interactions with the conversational AI. However, this approach encounters challenges when dealing with unseen interactions, which refers to new user interactions not covered by the user history index. This paper introduces our “Collaborative Query Rewriting” approach, which utilizes underlying topological information to assist in rewriting defective queries arising from unseen interactions. This approach begins by constructing a “User Feedback Interaction Graph” (FIG) using historical user-entity interactions. Subsequently, we traverse through the graph edges to establish an enhanced user index, referred to as the “collaborative user index”. We then delve deeper into the utilization of Large Language Models (LLMs) to assist in graph construction by understanding user preferences, leading to a significant increase in index coverage for unseen interactions. The effectiveness of our proposed approach has been proven through experiments on a large-scale real-world dataset and online A/B experiments.