Every Bite Is an Experience: Key Point Analysis of Business Reviews

Proceedings of the 59th Annual Meeting of the Association for Computational Linguistics and the 11th International Joint Conference on Natural Language Processing, pages 3376–3386 August 1–6, 2021. ©2021 Association for Computational Linguistics 3376 Every Bite Is an Experience: Key Point Analysis of Business Reviews Roy Bar-Haim, Lilach Eden, Yoav Kantor∗, Roni Friedman, Noam Slonim IBM Research {roybar,lilache,yoavka,roni.friedman-melamed,noams}@il.ibm.com Abstract Previous work on review summarization fo- cused on measuring the sentiment toward the main aspects of the reviewed product or busi- ness, or on creating a textual summary. These approaches provide only a partial view of the data: aspect-based sentiment summaries lack sufficient explanation or justification for the aspect rating, while textual summaries do not quantify the significance of each element, and are not well-suited for representing conflict- ing views. Recently, Key Point Analysis (KPA) has been proposed as a summarization frame- work that provides both textual and quantita- tive summary of the main points in the data. We adapt KPA to review data by introduc- ing Collective Key Point Mining for better key point extraction; integrating sentiment analy- sis into KPA; identifying good key point can- didates for review summaries; and leverag- ing the massive amount of available reviews and their metadata. We show empirically that these novel extensions of KPA substantially improve its performance. We demonstrate that promising results can be achieved without any domain-specific annotation, while human su- pervision can lead to further improvement. 1 Introduction With their ever growing prevalence, online opinions and reviews have become essential for our every- day decision making. We turn to the wisdom of the crowd before buying a new laptop, choosing a restaurant or planning our next vacation. However, this abundance is often overwhelming: reading hun- dreds or thousands of reviews on a certain busi- ness or product is impractical, and users typically have to rely on aggregated numeric ratings, com- plemented by reading a small sample of reviews, which may not be representative. The vast majority of available information is left unexploited. ∗First three authors equally contributed to this work. Opinion summarization is a long-standing chal- lenge, which has attracted a lot of research interest over the past two decades. Early works (Hu and Liu, 2004; Gamon et al., 2005; Snyder and Barzi- lay, 2007; Blair-goldensohn et al., 2008; Titov and McDonald, 2008) aimed to extract, aggregate and quantify the sentiment toward the main aspects or features of the reviewed entity (e.g., food, price, service, and ambience for restaurants). Such aspect- based sentiment summaries provide a high-level, quantitative view of the summarized opinions, but lack explanations and justifications for the assigned scores (Ganesan et al., 2010). An alternative line of work casts this problem as multi-document summarization, aiming to create a textual summary from the input reviews (Carenini et al., 2006; Ganesan et al., 2010; Chu and Liu, 2019; Braˇzinskas et al., 2020b). While such sum- maries provide more detail, they lack a quantitative view of the data. The salience of each element in the summary is not indicated, making it …

Content truncated to 500 words