Aspect-Adaptive Knowledge-based Opinion Summarization

Guan Wang, Weihua Li, Edmund M.-K. Lai, Quan Bai

Abstract: The increase in online information has overwhelmed users with opinions and comments on various products and services, making decision-making a daunting task. Text summarization can help by distilling long or multiple documents into concise, relevant content. Recent advances in Large Language Models (LLM) have shown great potential in this area. The existing text summarization approaches often lack the “adaptive” nature required to capture diverse aspects in opinion summarization, which is particularly detrimental to users with specific preferences. In this paper, we introduce an Aspect-adaptive Knowledge-based Opinion Summarization model for product reviews. This model generates summaries that highlight specific aspects of reviews, providing users with targeted, relevant information quickly. Our extensive experiments with real-world datasets explicitly demonstrate that our model surpasses current state-of-the-art methods. It effectively adapts to user needs, producing efficient, aspect-focused summaries that help users make informed decisions based on their unique preferences.