图书情报知识 ›› 2021, Vol. 38 ›› Issue (6): 60-72.doi: 10.13366/j.dik.2021.06.060

• 学术聚焦·细粒度引文分析 • 上一篇    下一篇

引文情感分析方法研究综述

董克,吴佳纯   

  • 出版日期:2021-11-10 发布日期:2022-01-13

Literature Review on Citation Sentiment Analysis Methods

  • Online:2021-11-10 Published:2022-01-13

摘要: [目的/意义]总结引文情感分析方法的发展脉络,剖析引文情感分析方法面临的挑战与未来发展方向。[研究设计/方法] 回顾了国内外引文情感分析相关的研究成果,归纳了引文情感分析的主要流程包括引文内容抽取、引文情感分类、引文情感识别;系统分析了相关研究成果,总结了引文情感分析方法在学术评价、信息检索等领域的应用。[结论/发现] 引文情感分析方法在技术与应用上已经取得了丰富的研究成果;其在引文上下文抽取、语料标注和特征工程中依旧面临许多问题与挑战;端到端引文情感分析、多特征引文情感分析、多模态引文情感分析以及基于时间序列的引文情感分析是未来发展的重要方向。[创新/价值]揭示了引文情感分析方法的发展进程与特点,阐述了引文情感分析方法的未来发展方向。

关键词: 引文情感分析, 引文内容, 情感极性, 情感分类, 学术评价

Abstract: [Purpose/Significance]This article summarizes the development of citation sentiment analysis. It also analyzes the challenges faced by citation sentiment analysis and its future direction.[Design/Methodology]By reviewing the results of relevant research on citation emotion analysis at home and abroad, this article put forward the framework of citation sentiment analysis including citation content extraction, citation sentiment classification, and citation sentiment recognition system; It systematically analyzed the existing related research results, and the applications of citation sentiment analysis in academic evaluation, information retrieval and other fields were also summarized.[Findings/Conclusion]Citation sentiment analysis has already made abundant important achievements in technologies and applications. However, it still faces many problems and challenges, such as citation context extraction, corpus annotation and feature engineering; At the same time, end-to-end citation sentiment analysis, multi-feature citation sentiment analysis, multi-modal citation sentiment analysis, and citation sentiment analysis based upon time series are the significant future development directions.[Originality/Value]This article reveals the development progress and characteristics of citation sentiment analysis, and expounds its future development directions.

Key words: Citation sentiment analysis, Citation context, Emotion polarity, Emotion classification, Academic evaluation