{"title":"Unsupervised extractive opinion summarization based on text simplification and sentiment guidance","authors":"Rui Wang , Tian Lan , Zufeng Wu , Leyuan Liu","doi":"10.1016/j.eswa.2025.126760","DOIUrl":null,"url":null,"abstract":"<div><div>Unsupervised opinion summarization aims to extract representative content from a set of reviews without relying on golden references. Traditional unsupervised methods often struggle with non-consensus opinions and lack conciseness in the extracted summaries, which may prevent users from making swift and informed decisions. To tackle these challenges, we propose a novel two-stage unsupervised opinion summarization method based on text simplification and sentiment guidance. In the first stage, we leverage a pre-trained language model to simplify complex sentences into concise and clear forms. In the second stage, our method identifies and clusters sentences based on sentiment information. Summary sentences are subsequently extracted to align with the overall sentiment tendency, ensuring consistency and representativeness. Experimental results on the SPACE and AMAZON benchmark datasets demonstrate performance improvements, confirming the efficacy of our approach in addressing the identified challenges.</div></div>","PeriodicalId":50461,"journal":{"name":"Expert Systems with Applications","volume":"272 ","pages":"Article 126760"},"PeriodicalIF":7.5000,"publicationDate":"2025-02-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Expert Systems with Applications","FirstCategoryId":"94","ListUrlMain":"https://www.sciencedirect.com/science/article/pii/S0957417425003823","RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q1","JCRName":"COMPUTER SCIENCE, ARTIFICIAL INTELLIGENCE","Score":null,"Total":0}
引用次数: 0
Abstract
Unsupervised opinion summarization aims to extract representative content from a set of reviews without relying on golden references. Traditional unsupervised methods often struggle with non-consensus opinions and lack conciseness in the extracted summaries, which may prevent users from making swift and informed decisions. To tackle these challenges, we propose a novel two-stage unsupervised opinion summarization method based on text simplification and sentiment guidance. In the first stage, we leverage a pre-trained language model to simplify complex sentences into concise and clear forms. In the second stage, our method identifies and clusters sentences based on sentiment information. Summary sentences are subsequently extracted to align with the overall sentiment tendency, ensuring consistency and representativeness. Experimental results on the SPACE and AMAZON benchmark datasets demonstrate performance improvements, confirming the efficacy of our approach in addressing the identified challenges.
期刊介绍:
Expert Systems With Applications is an international journal dedicated to the exchange of information on expert and intelligent systems used globally in industry, government, and universities. The journal emphasizes original papers covering the design, development, testing, implementation, and management of these systems, offering practical guidelines. It spans various sectors such as finance, engineering, marketing, law, project management, information management, medicine, and more. The journal also welcomes papers on multi-agent systems, knowledge management, neural networks, knowledge discovery, data mining, and other related areas, excluding applications to military/defense systems.