论文部分内容阅读
【目的/意义】明确用户评论的情感倾向及其主要影响因素,有助于政府了解网民的态度及观点,更好地引导和控制舆情发展。【方法/过程】运用八爪鱼软件采集新浪政务微博上的辟谣信息,获取数据75847条,构建多元Logistic回归模型对用户情感倾向的影响因素进行分析。【结果/结论】研究发现:用户对政府辟谣信息表现出不同的情感倾向,其中,中性情感倾向用户比重最小,消极情感用户和积极情感用户占比接近;内容特征变量、文本特征变量、用户特征变量均对用户情感倾向有显著影响,但影响方式各异。
[Purpose / Significance] Defining the emotional bias of user comments and its main influencing factors will help the government understand the attitudes and opinions of Internet users and better guide and control the development of public opinion. 【Method / Procedure】 Using the octopus software to collect the rumor information from Sina Weibo and get 75847 data, and construct multivariate Logistic regression model to analyze the influencing factors of users’ affective tendencies. [Results / Conclusion] The study found that users showed different affective tendencies toward the rumor information of the government, among which the proportion of neutral affective tendencies was the smallest, the proportion of negative affective users and positive affective users was close to each other; content feature variables, text characteristic variables, The characteristic variables all have a significant impact on the user’s affective tendencies, but the influencing ways are different.