Study On Weibo Hot Event Data Mining Based On Spatial-temporal Distribution Analysis And User Relationship Analysis

In recent years,vicious affairs that caused by Weibo have created the extremely bad social impact,seriously influenced the stability of the society.This study focuses on the mining technologies of hot event propagation.Literature review shows that previous studies about network hot events are mainly about description analysis of influential factors on each propagation step,few studies about logistics case study,and many analysis methods are qualitative,few are quantitative.For this reason,this paper takes the “Tianjin Tanggu explosion” and “contraceptives eel” as the empirical study object,using the social network analysis to explore the quantitative methods for Sina Weibo hot events of the overall network structure identification,subgroup characteristics analysis and key nodes mining.First,using python programming to obtain data from Sina Weibo web pages,after processing the original data to get the user relationship and spatial-temporal characteristics,finally,based on the mathematical statistics tools and UCINET software to visualize results.The analysis result showed that,the validity of social network analysis,natural language process methods,spatial-temporal analysis in this study is verified,and it also supplied a new idea for the research on Weibo hot event communication.