广电网络舆情分析架构与微博文本情感分析研究

Research on Public Opinion Analysis Framework of Broadcasting & TV Network and Sentiment Analysis of Weibo Text

网络舆情分析对于广播电视和网络视听传播影响力分析有着重要的作用和意义,情感分析是网络舆情研究中的重要组成部分。本文首先网络舆情和情感分析与广电突发事件间的关系,随后给出广电的网络舆情分析架构,在此架构下,提出一套基于BERT进行微博文本的情感分析方法,对涉及网络舆情情感的统计指标进行分析,并结合近期两个网络舆情实例,参考百度搜索指数与网页搜索结果进行深入分析,情感分析结果与上述参考数据基本一致。

Network public opinion analysis plays an important role and signifcance for infuence analysis of radio and television and network audio-visual communication. Sentiment analysis is an important part of network public opinion research. First of all, this paper analyzes the relationship between network public opinion, sentiment and radio and television emergencies. Then it gives analysis framework of broadcasting and television network public opinion. Under this framework, a set of sentiment analysis methods based on BERT for Weibo text is proposed, which analyzes statistical indicators related to network public sentiment. The paper also combines two recent examples, carries out in-depth analysis with reference to Baidu search index and web search results. Sentiment analysis results of this paper are basically consistent with the above reference data.