Font Size: a A A

Research On Multi-dimensional Evolution Analysis Method Of Public Security Events Web Public Opinions

Posted on:2020-12-13Degree:MasterType:Thesis
Country:ChinaCandidate:H HuFull Text:PDF
GTID:2416330590960219Subject:Safety engineering
Abstract/Summary:PDF Full Text Request
With the continuous development of Internet and mobile terminal technologies and the increasing number of mobile phone users,microblog,website comment columns,blog posts,etc.have become the main positions of online public opinion,and have had a significant impact on public opinion and government decision-making.The existing evolutionary analysis of the public security incident web public opinion is still insufficient.Therefore,research on the evolutionary analysis methods of public safety events and analyzing the evolution law of public security events web public opinion play an important role in government emergency management and decision-making.This paper makes an overall analysis of the evolution process of public security incident web public opinion,and studies the development of events from the public security event web public opinion,theme and public sentiment orientation.The dynamic characteristics of each stage of the public opinion event web public opinion evolution cycle are analyzed in detail.This paper analyzes microblog public opinion with the group social security incident of “Beijing Red,Yellow and Blue Child Abuse”.Aiming at the topic evolution of public security events in microblog,this paper proposes an MD-HDP topic model,which uses time features and topic tags to accurately extract relevant microblog texts,and then separately models and mines event texts in different time windows.The evolution of event themes on the time axis;for the public sentiment evolution of microblog public safety events,this paper proposes an improved PMI-IR algorithm,redefining the extraction mode of emotional phrases and the choice of standard words,and this paper proposes an emotional classification method that is more suitable for microblog comment texts,and mines the evolution characteristics of event emotions on the time axis.Finally,it explores the evolution law of events from the multiple dimensions of sensational heat evolution,event theme evolution and public sentiment tendency evolution.The department can grasp the development trend of the event more accurately,and intervene more reasonably at the time node to guide the direction of public opinion and avoid causing secondary harm.
Keywords/Search Tags:Public safety events, Topic mining, Sentiment analysis, Public opinion evolution
PDF Full Text Request
Related items