Design And Implementation Of Influence Measurement System Of Regional Microblogging Users

In recent years,"microblogging control" has become a buzzword on the Internet. Themicroblogging’s rapid development and its impact have set off a revolution of140characters. With the rapid development of Sina and Tencent microblogging system, it’sextensive influences on crackdown of child trafficking,7-23motor car accident and rightsafeguarding make people feel the great power of microblogging personally and make itsimpact on public opinion and information spreading, gain more and more publicattentions.User nodes with strong influences often play a key role in the appearance anddevelopment of a microblogging event. This thesis focused on the regional user ofmicroblogging, measure the influence of these users by the building of following networkand repost network. Our work includes the following points:1) locking of the regionaluser using their provincial characteristics,2) crawling users’ data on the microbloggingsystem with Open API,3) analyzing the two networks’ topology and basic attribute usingusers data, and4) measuring user’s influence by introducing and improving the PageRankalgorithm.Finally, the following results are concluded: there are no coordinate nodes inmicroblogging repost behavior with31.4%users accounting for80%reposts. This showsthat a small amount of users drive the media in microblogging. Consider the unevenreposts based on the PageRank algorithm, we use uneven votes to complete the iteration,which is fit close to the actual results of influences. In addition, the results prove that thenumber of user’s fans is not proportional to its influence.