Research On Event Association Analysis And Visualization Based On Text Mining

Due to the rapid development of the Internet,the network resources show explosive growth,and the event text data is also increasing exponentially.There are multiple and complex relationships among events.Analyzing the relationships correctly can make people understand the events more quickly and clearly,so how to get the relationships in a large number of data becomes an urgent problem.In the process of event correlation analysis,due to the uncertainty,instantaneous change and other reasons of network resources,it is necessary to select data with complete data,rich events,stable and controllable data for analysis.Therefore,this paper takes famous works as the research carrier,trying to complete the research of text event correlation analysis,providing certain social reference value for other work.The main work and innovations of this paper are as follows:(1)Data preprocessing: the text data is segmented,the stop words and part of speech are removed,and the segmentation process is optimized,so that the accuracy of segmentation is increased from 48.3% to 78.9%.(2)Event extraction and relevance analysis: in the event extraction stage,select the named entity and core trigger words of the event,and then conduct relevance analysis for the extracted elements,mainly including the constituent relationship of the event,the connection of the core verbs and the overall analysis of the event text.(3)Data visualization analysis: using the construction of knowledge map,through the data visualization tools for specific fields to propose relevant solutions,and the event analysis results are clearly displayed in the platform.