Thedesign And Implementation Of Analysis System Of Micro-blog Public Opinion Information Dissemination

With the progress of technology,the impact of the Internet on human life has been continuously strengthened.More and more people can express their ideas in political,economic,commercial and other fields through the social media such as micro-blog,forum and blog.From the annual report of china internet public Opinion in 2017,we can find that microblog has become the second largest source of public opinion after traditional media.Microblog users mainly participate in interaction through comments,likes,forwarding and other behaviors.Actions such as intensification of negative public sentiment,over-distortion of opinion leaders and insufficient motivation of mainstream media will have an impact on the maintenance of stability.Because the network public opinion and big data are highly consistent,the traditional information processing technology can not effectively analyze the large-scale public opinion network data,how to quickly and accurately analyze the valuable information in the public opinion network is very important.The dissemination and analysis system of public opinion information on microblog is designed and implemented in this paper to analyze and grasp the evolution law of public opinion clearly and intuitively.The system collects information of network public opinion events by simulating the login of Sina Blog,then it constructs public opinion network and identifies key nodes and propagation path of the network by visualization technology.It restores the process of microblog event diffusion,and provides a good interactive public opinion analysis tool for relevant personnel.The main research findings of this paper are as follows:(1)It uses node.js to simulate the login of Sina Blog,automatically crawls to microblog and forwards web page information through the web crawler,then it writes the acquired data into the database and constructs a network map of microblog public opinion forwarding based on data from the database.In addition,it can analyze key nodes and propagation paths in public opinion network graph by specific algorithms.(2)A new method called ActiveRank for identifying key nodes is proposed according to the active value of users and the structure and link relationship of microblog public opinion network.This method divides microblog users into three categories: ordinary users,active users and opinion leaders to define user activity values.Then it finds out opinion leaders in the public opinion dissemination network by combining activity value with PageRank algorithm for identifying key nodes in social networks(3)The analysis results of microblog forwarding on the market are mostly displayed in the form of charts.In this paper,the public opinion network chart is used to visually present the process of microblog forwarding.The paper uses D3.js' s force directed graph to show the network of public opinion dissemination,and considers forwarding time in visual display to grasp the change of microblog forwarding network in time as a whole.