摘要
基于话题与时间序列分析揭示公共安全事件衍生舆情的形成和演化规律,有助于政府部门预警、研判与治理衍生舆情。以与"长春长生疫苗事件"有关的微博数据为研究样本,通过构建话题发现模型,识别衍生舆情,利用ARIMA时间序列模型来拟合和预测衍生舆情的形成与演化,可以发现:信息不对称现象使得公共安全事件常在蔓延期与衰退期产生衍生舆情,尤其是当原生舆情处理不当,或者是涉及弱势群体、政府官员等争议性话题时;衍生舆情与政府举措、实体遭遇等重要时间节点事件密切相关。因此,有关部门需加强各方沟通,恰当处理原生舆情及应对措施,尤其是在蔓延期与衰退期需提前预防衍生舆情爆发。
In order to assist government departments in early warning,research and governance of derived public opinion,an analysis based on topic and time series is used to reveal the formation and evolution of public opinion derived from public security events.Taking the microblogging related to the"Changchun Changsheng vaccine event"as an example,we build a topic discovery model,identify derived public opinion,and use ARIMA time series model to fit and predict the formation and evolution of derived public opinion.It is found that information asymmetry generally derives public opinion in the spread and decline of public security events,especially when original public opinion is not properly handled,or involves controversial topics such as vulnerable groups,government officials,etc.Derivative public opinion is closely related to events on important time nodes such as government actions and entity encounters.Therefore,departments concerned should strengthen mutual communication,properly handle original public opinion and take response measures to prevent the outbreak of derivative public opinion in advance,especially in the spread and decline of public security events.
作者
安璐
代园园
周亦文
AN Lu;DAI Yuanyuan;ZHOU Yiwen
出处
《公安学研究》
2020年第1期14-31,123,共19页
Journal of Public Security Science
基金
教育部哲学社会科学研究重大课题攻关项目“提高反恐怖主义情报信息工作能力对策研究”(17JZD034)
国家自然科学基金重大课题“国家安全大数据综合信息集成与分析方法”(71790612)
国家自然科学基金青年项目“突发公共卫生事件社交媒体信息主题演化与影响力建模”(71603189)。
