摘要
笔者融合多个词典建立情感词典,根据建立的规则计算文本情感极性,并对不含情感词的语句情感值计算进行了讨论,最终得到一套基于情感词典的情感倾向性分析方法,实现了文本的情感归类。在分析文本词频、情感比例及其时序变化的基础上,探究评论舆情导向和时序发展特征。使用微博评论数据进行实验,发现整体上正面情感占主导地位,网民情绪呈现随着事件发布短时间波动后逐渐趋于稳定的特点。
This paper integrates multiple dictionaries to build an emotional dictionary,calculates the text-emotional polarity according to established rules,and discusses the emotional value of sentences without emotional words,finally obtains a set of emotional tendency analysis methods based on the dictionary,which realizes text sentiment classification.Based on the word frequency,sentiment ratio and time series,we explore the comment orientation and time series development.Experiment using Weibo comment data reveals that positive emotions dominate as a whole,and netizens'emotions tend to stabilize gradually after a period of fluctuations following the release of events.
作者
纪佳昕
JI Jiaxin(Information Engineering University Luoyang Campus,Luoyang Henan 471003,China)
出处
《信息与电脑》
2021年第11期33-35,共3页
Information & Computer
关键词
词典
情感分析
微博评论
舆情
dictionary
sentiment analysis
Weibo comments
public opinions
