Research And Analysis Of Network Emergency Based On Micro-Blog

The Internet has gradually developed into the mainstream remarks platform nowadays, at the same time, the rise of a kind of self-media-micro-blogs has brought an unprecedented situation for the dissemination of information. Information started to become multi-sources that anyone can become an individual post. Information dissemination started to become faster speed, the viral spread allows a hot topic to spread throughout the network within4hours. Information started to become more interactive, content of it becomes richer, distance becomes closer to the event center. All of these changes are brought by networks and social media.These changes also brought great significance on the research of network emergency. There are an unprecedented surge amount of information, newly generated information on the micro-blog platform are hundreds and billions which made analysis more difficult. At the same time, however, the micro-blog platform is a highly characteristic platform, which made the ways of how the topics are polymerized, forwarded and disseminated becomes able to track.In this paper, through the study of information dissemination and propagation characteristics on micro-blog platform, we brought out a way to explore and track hot topics. First of all, we filter the information by their initial property, extracts about1.94%of the data only. Secondly, based on the polymerization topic in the heat-factor, we extracted information sources that can probably develop into unexpected event. Finally, we extract information according to the feature of the micro-blog itself, get those that are most likeable to cause large social influence for further tracking and study. Experiment shows that the proposed method can extracte micro-blog information effectively from the massive amout generated daily. It can reduce about5.81million newly send message to around500key messages to track, making select rate lower than0.01%. Micro-blogs that are extracted shows a wider spreading potential, proves that the micro-blogs extracted do have the possibility to become hot topics.