Life Cycle Identification and Analysis of Microblog hot Topics

Through the life cycle identification and feature analysis of hot topics, we can monitor the influence, spreading time and the development trend of it. So this technology provides support for monitoring the spread of network information, and is an important topic for early warning of Internet public opinion. [Method/Process] Based on the hot topic of microblog, this paper firstly constructs one hot topic by analyzing the different themes at each stage, the life intensity index and trend. Then, classified 75 to-pics that collected previously into five categories by SOM clustering algorithm. Finally, describe the characteristics of five life cycle curves through the macro-microscopic analysis. [Result/Conclusion] Empirical analysis shows that hot topics have features like short cycle, rapid growth, rapid decline, short maturity, and the vast majority of hot topics only have one peak, but not more than two peaks. At the same life cycle stage of the five curves, we find the life intensity index has some features in common, and put forward two indicators to describe the features. The result can be used to monitor the hot topics on internet.