Font Size: a A A

Design And Implementation Of Data Analysis System For Government Microblog

Posted on:2021-02-02Degree:MasterType:Thesis
Country:ChinaCandidate:G J WuFull Text:PDF
GTID:2416330623458507Subject:Engineering
Abstract/Summary:PDF Full Text Request
Nowadays,government microblog plays an increasingly important role in government work.The content of government microblog is various,and the number of daily releases is increasing.There is an urgent need to improve the management and operation efficiency of government microblog,grasp the public opinion tendency of recent government microblog and guide the public opinion in a timely manner,and get the government affairs topics that the public is most concerned about.The research in the field of data analysis of government microblog involved in this paper has a good application prospect for solving the problems of government microblog operation and management.Through investigation and research on the development status of government microblog in recent years,this paper found that government microblog operators could not grasp the hot topics concerned by the public inside government microblog,could not timely understand the public opinion and emotional tendency of government microblog information,and had low efficiency in government microblog operation and management.In addition,the amount of data in the government microblog account keeps increasing,with complex data features and hidden potential values.According to the urgent needs of government microblog,this paper designs and implements a data analysis system for government microblog.This system has realized the functions of extracting the topics of government microblog that the public cares most,obtaining the public's emotional tendency towards the recent popular microblog,analyzing the user behavior and user characteristic information of microblog,and collecting the hot release material information regularly.Taking the government microblog account "Shanghai release" as an example.The whole system is divided into three subsystems,namely,data collection subsystem,data analysis subsystem and data visualization subsystem.The data acquisition subsystem utilizes a distributed crawler Scrapy framework to collect data,the data analysis subsystem uses Spark distributed system for data analysis,including Chinese word segmentation,feature extraction,LDA topic extraction,emotion analysis,SparkSQL data analysis,etc,the data visualization subsystem uses SSM and Echarts framework to visualize the analyzed data.In this paper,an HRCA model(the heat value of microblog affairs)is proposed for the analysis of government microblog,based on the analysis of users' interest in microblog forwarding,comments and thumb up,as well as the potential popularity of users,the model calculates the microblog content that microblog users attach great importance to.The model is combined with LDA algorithm to extract hot topics on government microblog.At the same time,An EDS model(the emotional value of microblog affairs)is proposed to calculate the emotional value of government microblog comments.This model divides the comment text of government microblog into three sentence types: declarative sentence,exclamatory sentence and interrogative sentence,combined with the length of comment text,the affective dictionary is used to calculate the affective value of text comment.A set of data analysis system for government microblog is developed.The system solves the problems of slow traditional data analysis,insufficient sample data,inaccurate information,fuzzy hot topics of government microblog,inability to timely obtain public opinion tendency,difficulty in analyzing user behavior and characteristic data,and tedious collection of publishing materials,and improves the operating efficiency of government workers for government microblog.
Keywords/Search Tags:government microblog, text mining, emotion analysis, HRCA model, EDS model, visualization
PDF Full Text Request
Related items