Font Size: a A A

Clustering-based Acquisition Method Researching Of Microblog Public Opinion

Posted on:2016-12-11Degree:MasterType:Thesis
Country:ChinaCandidate:R LiuFull Text:PDF
GTID:2348330542475375Subject:Computer application technology
Abstract/Summary:PDF Full Text Request
With the rapid development of Internet,the application of social network has achieved the explosive growth.Currently,the Chinese microblog platforms have become one of the most important applications of the Chinese social networks,which have aggregate the huge amount of network public opinion at the same time.The rapid and accurate achievement of network public opinion caused by microblog information is able to help administrative department and network carrier to more efficiently manage the microblog platform and purify the network environment.The microblog information has the characteristics of non-structured data and the huge transmission quantity,which enable the traditional methods of data mining not to high-efficiently and accurately achieve the microblog public opinion.How to rapidly and accurately achieve the network public opinion produced by microblog information has become a new research direction in data mining field.To achieve the network public opinion by microblog information need to firstly achieve the relevant structural characteristics besides the microblog information ontology,namely,web mining which is able to convert the microblog information,the related information relied by microblog and the information of microblog owner into characteristic vector needed by clustering analysis,then enable the text mining of the microblog information and enable the data mining combining with the data produced by Web mining after extracting the characteristic vector.The previous algorithms usually ignored the structural characteristics besides the microblog information ontology and just analyzed the text from the extraction of microblog information,which caused it difficult to accurately classify the public opinion in the process of clustering analysis.This paper introduces the pagerank algorithm which is originally applied into network page into the clustering analysis of microblog information,and then proposes the clustering algorithm of the weighted microblog information,which simplifies the calculation and improve the accuracy of clustering calculation,then achieve the acquiring method of microblog information based on clustering analysis.
Keywords/Search Tags:Microblog information, network public opinion, data mining, Cluster analysis, weighted objects
PDF Full Text Request
Related items