Improving User Experience with Personalized Review Ranking and Summarization

TL;DR

This work tackles information overload in online consumer reviews by proposing a personalized framework that jointly ranks reviews and generates abstractive summaries aligned with individual user preferences. It fuses sentiment signals from both star ratings and textual content with user preference profiles derived from historical reviews, using a semantic scoring function to rank unseen reviews. Top-ranked reviews are then distilled into personalized summaries via an LLM, enabling faster, more relevant decision-making. In a study with 70 participants across multiple product categories, the approach improved satisfaction, perceived relevance, and decision-making confidence while reducing reading time, highlighting practical benefits for scalable, user-centric review platforms. The method offers a scalable path to more adaptive decision support in information-rich e-commerce contexts, with potential extensions to broader domains and multimodal data.

Abstract

Online consumer reviews play a crucial role in guiding purchase decisions by offering insights into product quality, usability, and performance. However, the increasing volume of user-generated reviews has led to information overload, making it difficult for consumers to identify content that aligns with their specific preferences. Existing review ranking systems typically rely on metrics such as helpfulness votes, star ratings, and recency, but these fail to capture individual user interests and often treat textual sentiment and rating signals separately. This research addresses these limitations by proposing a personalized framework that integrates review ranking and abstractive summarization to enhance decision-making efficiency. The proposed system begins by modeling each user's sentiment through a hybrid analysis of star ratings and review content. Simultaneously, user preferences were derived from historical reviews using sentence embeddings and clustering, forming semantic profiles aligned with thematic and sentiment dimensions. A relevance scoring algorithm matched these profiles with unseen reviews based on sentiment and aspect similarity. Top-matched reviews were then summarized to reflect individual interests. A user study with 70 participants demonstrated that the personalized approach improved satisfaction, perceived relevance, and decision-making confidence, while reducing time spent reading. The results highlight the method's effectiveness in alleviating information overload and delivering content tailored to user-specific preferences, emphasizing its value in enhancing user experience in review-rich decision-making environments.

Figures (14)

• Figure 1: Research Methodology
• Figure 2: LLM Prompt Used to Generate Summary of The Reviews
• Figure 3: LLM Generated Personalized Summary of the Ranked reviews
• Figure 4: Signup Interface
• Figure 5: Product Rate and Review Interface