Hotspot Prediction And Analysis Of Propagating Of Topics Based On Microblog

Microblog, as a new social network, has attracted a large number of users, benefitting from effective information dissemination, content diversity, wide coverage. Microblog is playing a more and more important role in spread of public opinion. The contents of many microblogs are around the same event, which constitute a topic. The hotspot prediction and propagating analysis of topics could help people keep up with the latest trend about events, which can be used in the mining of news and the regulation of the bad information in advance.In this paper, we introduce a complete system, including data acquisition and preprocessing, topic clustering and hotspot prediction.The main works of this paper are:1. As we know, a lot of information is meaningless on the microblog, such as advertising, business promotion and publicity activities. In the preprocessing stage, features from two aspects of users and microblog content are extracted, and then information filtering through the SVM classifier is realized.2. In order to predict the hot topic, we divide the propagating process into three stages:incubation, outbreak and recession period, and analyze the related features in each period in detail. The experiments show that the good results using the four features to describe the propagation:proportion of VIP, distribution of the fans number, distribution of topic flags and external links.3. A classification method based on a latent source model is proposed, which uses the timing signal to represent the trends of topics. The result shows that, we can detect trends before Sina does83.5%of the time, with a mean value of1.6hour in advance.