Abstract:

Combined with time series analysis, cluster analysis and complex network analysis, this paper analyzes user classification and topic evolution in social blog. Firstly, according to the user activity in the social media, the authors set the threshold of relationship intensity and extract the core users. Secondly, this paper does the statistics of the annual hot topics concerned by the whole community. Finally，with comparing to hot topics tagged by core users, empirical analysis shows that the core users has significant impact on the derivative of hot topics and noncore users should not be ignored for the long tail effect.

Key words: Social blog;   , Topic evolution;     , MetaFilter;    , User classification