A hot topic phrase selection based on LDA for Chinese tweets
PU Jian-zhong · Luoyang Shifan Xueyuan xuebao · 2012
This paper proposes a method that combines LDA model and related techniques of the natural language information processing in a bid to extract hot topic phrases in Chinese microblogs in Twitter.To corroborate the effectiveness of this method,this paper selects 20,923 pieces of Chinese Tweets as samples.The extracted topic phrases are generally within expectation,demonstrating that this model is capable of recognizing hot topics and describing them.