基于情感特征词的网络舆情参与者情感特征研究

Research on the Emotional Characteristics of Internet Public Opinion Participants Based on Emotional Feature Words

以近年来国内学术不端事件为切入口,首先,梳理短文本情感倾向性分析的基本过程和方法,阐述基于情感特征词并融合情感贡献度的情感倾向性计算方法;然后,利用爬虫工具爬取了12个学术不端事件的微博评论语料库;最后,构建出一个短文本情感倾向性分析可视化系统,并利用该系统对前期采集的语料库进行分析。分析结果全方位、形象化地展示三类典型事件中网民的情感倾向及其情感表达特点,为相关部门更有效地干预和引导舆论方向、提出应对和管理类似事件的策略提供参考。

Based on the domestic academic misconduct events in recent years,firstly,this study combs the basic process and method of sentiment orientation analysis of short text,and expounds the calculation method of emotional orientation based on emotional feature words and emotional contribution;Then,using the crawler tool to crawl the microblog comment corpus of 12 academic misconduct events;Finally,a visual system for sentiment orientation analysis of short texts is constructed,and the corpus collected in the early stage is analyzed by the system.The analysis results can show the emotional tendency and emotional expression characteristics of netizens in three typical events in an all-round and visualized way,which provides a guarantee for relevant departments to grasp the focus of public opinion and development direction of specific events more accurately,and puts forward strategies to deal with and manage similar events.