摘要
[目的/意义]情感分析技术广泛应用于网络舆情方面,该技术可以有效地区别分析网民在网络社交平台所发布言论的情感极性。受到应用领域的限制,基础的情感词典并不能满足特定应用领域对于情感分析的需求。本文构建的词典可以满足网络舆情领域的情感分析需求。[方法/过程]先使用TF-IDF和TextRank提取种子词,然后采用SO-PMI算法构建突发事件网络舆情领域的情感词典。[结果/结论]使用该情感词典对“昆山反杀案”这一突发事件的微博评论进行情感分析,证明了所构建的情感词典在一定程度上具有准确性及适用性。
[Purpose/significance] Sentiment analysis technology is widely used in network public opinion,which can effectively distinguish and analyze the emotional polarity of netizens’ comments on network social platforms. Limited by the application field,the basic emotional dictionary cannot meet the needs of sentiment analysis in a specific application field. The dictionary constructed in the paper can meet the needs of sentiment analysis in the field of network public opinions. [Method/process] The paper uses TF-IDF and TextRank to extract seed words,and uses SO-PMI algorithm to build an emotional dictionary in the field of emergency network public opinion. [Result/conclusion] Using the emotional dictionary to analyze Weibo comments on the Kunshan Anti-Killing Case,it proves that the constructed emotional dictionary is accurate and applicable to a certain extent.
作者
管雨翔
王娟
刘静
秦瑞青
张鹏
Guan Yuxiang;Wang Juan;Liu Jing;Qin Ruiqing;Zhang Peng(Network Public Opinion Research Center of People’s Police University of China,Langfang Hebei 065000;School of Liberal Arts and Sciences North China Institute of Aerospace Engineering,Langfang Hebei 065000)
出处
《情报探索》
2023年第2期1-8,共8页
Information Research
基金
教育部人文社会科学青年基金项目“基于舆情大数据的突发事件网民情感风险感知与治理研究”(项目编号:20YJC630145)
河北省高等学校人文社会科学研究项目资助“大数据背景下高校网络舆情风险预警与应对策略研究”(项目编号:SQ2022026)
河北省高等学校人文社会科学研究项目“基于舆情大数据的群体性事件感知与应对策略研究”(项目编号:BJ2020210)
廊坊市哲学社会科学研究课题“面向政治安全的京津冀地区网络舆情风险研判与治理研究”(项目编号:2022013)成果。
