基于VAR模型的微博用户转发行为研究  

Research on the Retweeting Behavior of Weibo Users Based on VAR Model

作　　者：黄川林[1] 鲁艳霞[1,2] 崔艳清[1] HUANG Chuan-lin;LU Yan-xia;CUI Yan-qing(School of Information and Business Management,Dalian Neusoft University of Information,Dalian 116023,China;School of Management and Economics,Dalian University of Technology,Dalian 116023,China)

机构地区：[1]大连东软信息学院信息与商务管理学院,辽宁大连116023 [2]大连理工大学经济管理学院,辽宁大连116023

出　　处：《情报科学》2020年第5期148-153,共6页Information Science

基　　金：辽宁省社会科学规划基金项目“新时代网络舆情的大数据技术应用与研究”(L18BXW001);大连市社会科学项目“突发事件舆情演化机理及创新社会治理体制研究”(2019dlskyb216)。

摘　　要：【目的/意义】随着移动互联网的发展,微博的普及进一步加速了社会突发事件的传播。转发作为最重要的用户信息行为,在很大程度上预示了网络舆情的发展趋势。但是,鲜有研究关注微博内容中的心理语言使用与转发行为的关系。本研究拓展了心理语言学在社会突发事件情境下的应用领域,为政府或企业应急管理部门有效引导网络舆情提供了实践启示。【方法/过程】本文以九寨沟地震事件为例,基于LIWC文本分析工具研究了微博用户心理过程对于转发行为的影响,通过构建VAR向量自回归模型并进行格兰杰因果检验,确定了微博转发行为的心理语言影响因素,并进一步运用脉冲响应函数对转发行为进行了动态分析。【结果/结论】根据实证研究的结果,社会过程词和情感历程词对微博用户的转发行为具有一定的预测作用。【Purpose/significance】With the development of mobile Internet,The popularity of Weibo has further accelerated the spread of social emergencies.As the most important user information behavior,retweeting indicates the development trend of network public opinion.However,few studies have focused on the relationship between psycholinguistic use in We ibo content and retweeting behavior.This study expands the application field of psycholinguistics in social emergencies and provides practical enlightenment for the government or enterprise emergency management department to effectively guide online public opinion.【Method/process】Taking Jiuzhaigou earthquake as an example,based on the LIWC text analysis tool,this paper studies the influence of Weibo users'psychological process on retweeting behavior.By constructing VAR vector autoregression model and Granger causality test,the psycholinguistic influencing factors of Weibo retweeting behavior are determined and the dynamic analysis of retweeting behavior is further carried out by using impulse response function.【Re sult/conclusion】According to the results of empirical research,social process words and emotional process words can pre dict the retweeting behavior of Weibo users.

关 键 词：转发行为 心理语言 VAR模型 LIWC 

分 类 号：G206.3[文化科学—传播学] G252.0