基于主题挖掘技术的文本情感分析综述

Review of Text Emotion Analysis Based on Topic Mining Technology

[目的/意义]在互联网数据呈爆炸式增长的今天,如何提高情感分析的效率和准确性,从中挖掘更细粒度、更深层次的情感内涵成为学者们研究的热点。随着主题挖掘技术在处理大规模文本上所表现出的独特优势,越来越多的学者开始倾向于使用主题抽取技术去研究情感分析问题。鉴于此,文章对利用主题挖掘技术进行情感分析研究的相关文献进行分析和整理,为后续研究提供一定的借鉴价值。[方法/过程]围绕传统情感分析-主题挖掘技术-基于主题的文本情感分析这一主线展开,整理归纳国内外相关文献,并做必要评述。最后,对目前的研究困境和未来发展趋势进行总结与展望。[结果/结论]主要点明了传统情感分析中存在的问题和难点,据此提出将主题挖掘技术引入到情感分析研究中来,并将其划分为主题信息挖掘、情感倾向分析、情感演化分析和性能评价4项任务,对每一阶段任务的研究现状进行总结,为后续学者进行深入研究奠定理论基础。

[Purpose/significance]Nowadays,with the explosive growth of Internet data,how to improve the efficiency and accuracy of emotion analysis and dig out more fine-grained and deeper emotional connotation has become a hot research topic for scholars.With the unique advantages of topic-mining technology in processing large-scale texts,more and more scholars tend to use topic-extraction technology to study emotional analysis.In view of this,this paper analyzes and sorts out the relevant literature of emotional analysis research based on topic mining technology,so as to provide certain reference value for subsequent research.[Method/process]This paper focuses on the main line of traditional emotion analysis-topic mining technology-topic-based text emotion analysis,sorts out and summarizes relevant literature at home and abroad,and makes necessary comments.Finally,the paper summarizes the current research difficulties and the future development trend.[Result/conclusion]In this paper,we point out the problems and difficulties existing in the traditional emotion analysis,this paper introduce the topic mining technology to the study of emotion analysis,and can be divided into topic information mining,emotional tendency analysis,four task analysis and performance evaluation of emotional evolution,summarize the research status of each stage the task,and lay a theoretical foundation for subsequent scholars further study.