"It doesn't look good for a date": Transforming Critiques into Preferences for Conversational Recommendation Systems

TL;DR

This paper introduces a neural language model-based method to convert user critiques into positive preferences, enhancing conversational recommendation systems by better understanding and responding to user feedback.

Contribution

It presents a novel critique-to-preference transformation technique using large language models, improving recommendation relevance in conversational settings.

Findings

Critique-to-preference transformation improves recommendation quality.

Embedding matching and fine-tuning methods are effective for retrieval.

Three general cases explain the performance improvements.

Abstract

Conversations aimed at determining good recommendations are iterative in nature. People often express their preferences in terms of a critique of the current recommendation (e.g., "It doesn't look good for a date"), requiring some degree of common sense for a preference to be inferred. In this work, we present a method for transforming a user critique into a positive preference (e.g., "I prefer more romantic") in order to retrieve reviews pertaining to potentially better recommendations (e.g., "Perfect for a romantic dinner"). We leverage a large neural language model (LM) in a few-shot setting to perform critique-to-preference transformation, and we test two methods for retrieving recommendations: one that matches embeddings, and another that fine-tunes an LM for the task. We instantiate this approach in the restaurant domain and evaluate it using a new dataset of restaurant critiques. In an ablation study, we show that utilizing critique-to-preference transformation improves recommendations, and that there are at least three general cases that explain this improved performance.