内地青年网络社群涉台舆情的主题分布与情感特征——基于LDA模型和文本情感计算的大数据分析

Big Data Analysis of Taiwan-related Public Opinion in Youth Online Communities of Chinese Mainland:Based on LDA Model and Text Sentiment Calculation

综合使用LDA模型、文本情感计算等大数据方法与传统数据处理方法,分析内地主流青年网络社群中的舆情文本后发现,当前青年网络社群中的涉台议题表达可归入“政治”“文娱”“对抗”“生活”“交流”五大主题。在内容特征方面,“政治”和“交流”主题较易获得观看者的支持;“文娱”议题呈现“粉丝向”特征;“对抗”议题较易获得短期关注与转发;“生活”议题较易引发广泛讨论。在情感特征方面,涉台讨论呈现出“爱憎分明”的特点,最常被表达的情感是“赞扬”和“贬责”,但在积极向度的表达中,约有20%—25%属于揶揄和讽刺性用法。在互动形式方面,负面、强烈的情感表达较易引发关注和讨论。另外,用户性别与涉台讨论的情感向度显著相关。未来有必要进一步加强对青年网络社群的涉台舆情追踪,重视舆情引导,并强化青年网络社群中涉台交流的机制建设和积极情感表达。

This paper combines big data methods such as LDA modeling and text sentiment calculation with traditional data processing methods to analyze the public opinion texts in mainstream youth online communities in the mainland.It finds that the current expressions of Taiwan-related issues among these youth online communities can be categorized into five major topics:“politics”“entertainment”“confrontation”“life”and“communication”.In terms of content characteristics,the topics of“politics”and“communication”are more likely to gain viewer support;the“entertainment”topic shows a“fan-oriented”characteristic;the“confrontation”topic is more likely to receive short-term attention and retweets;while the“life”topic is more likely to generate extensive discussions.In terms of sentiment characteristics,Taiwan-related discussions are characterised by a“love-hate”dichotomy,with“praise”and“criticism”being the most frequently expressed sentiments.However,among the positive expressions,about 20%-25%are sarcastic or ironic.In terms of interaction forms,negative and intense expressions are more likely to generate attention and discussion.In addition,the gender of users is significantly correlated with the sentimentality of Taiwan-related discussions.In the future,it is necessary to further strengthen the tracking of Taiwan-related public opinion,pay attention to the guidance of public opinion,and consolidate the construction of mechanisms for Taiwan-related communication and positive emotional expressions within youth online communities.