A Mining System For Microblog Sensitive Information Based On Cloud Computing

With the rapid development of various information technologies,the Internet has rapidly become an essential part of our daily production and life.The rapid development and widespread population of the Internet make the people have many ways to access the things happened in nearby or in a faraway place.In general,the Internet improve the social reasonability and the awareness of protection rights of people.Sometimes,the internet users are more easily influenced or even misleaded by the viewpoints released on Internet and consequently form a bad network consensus,which endanger the social public property and people life.Therefore,the great energy of network consensus can not be ignored.Since microblog is one of the most influential network consensus media,the efficient monitoring of microblog consensus is more and more important.The consensus monitoring system right now partly solve sort of the problems of digging and tracking,but to sensitive information digging,its still low time efficiency,low accurate and high error rate.To solving these problems in the existed systems,the thesis provides a system,microblog sensitive information mining system based on cloud computing,for the increasing of users and influential of microblog.This system focuses on thunderbolt to build sensitive topic corpus,bringing in PageRank algorithm to handle society relations to get influence of users.Through natural language analysis and some other text analysis techniques,match high influential users' messages with the key words in corpus,then finds the sensitive people and render analysis result within web page so that user can check and handle at any time.This system has solved the problems of existing consensus monitoring system,monitor microblog users for the first time among this kind of system.Within private cloud platform built with MongoDB,high efficiency,stability will guarantee microblog the most efficient sensitive consensus monitoring.