政府辟谣信息的用户评论及其情感倾向的影响因素研究  

Research on User Reviews of Government Rumor-refuting Information and Factors Influencing Their Emotional Tendencies

作　　者：陈娟 刘燕平 邓胜利[2] 

机构地区：[1]华中农业大学公共管理学院信息管理系,湖北武汉430070 [2]武汉大学信息管理学院,湖北武汉430072

出　　处：《情报科学》2017年第12期61-65,72,共6页Information Science

基　　金：教育部人文社会科学重点研究基地重大项目(15JJD870001)

摘　　要：【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 ] The Identification of emotional tendencies of user reviews and their main influencing fac- tors has great significance for the government on understanding attitudes and views of network users, which can provide bet- ter guidance and controlling over the development of public opinion. [Method/process]75847 data associated with govern- ment rumor-refuting information have been collected by a collector named Bazhuayu from the government microblogs on Si- na website. The multiple Logistic regression model has been constructed to analyze the factors influencing users＇ emotional tendencies. [ Result/conclusion ]The results show that users show different emotional tendencies towards the government ru- mor-refuting information. The proportion of users with neutral emotional tendencies is the smallest, and that of users with negative or positive emotional tendencies is almost the same. All of the content characteristics, text characteristics, user characteristics have influences on user s＇ emotional tendencies. However, they share various ways.

关 键 词：辟谣信息 用户评论 情感倾向 多元LOGISTIC回归 影响因素 

分 类 号：G252.0[文化科学—图书馆学]