作者: Freddy Chong Tat Chua , William W. Cohen , Justin Betteridge , Ee-Peng Lim
关键词:
摘要: Many event monitoring systems rely on counting known keywords in streaming text data to detect sudden spikes frequency. But the dynamic and conversational nature of Twitter makes it hard select for monitoring. Here we consider a method automatically finding noun phrases (NPs) as Twitter. Finding NPs has two aspects, identifying boundaries subsequence words which represent NP, classifying NP specific broad category such politics, sports, etc. To classify an define feature vector using not just but also author's behavior social activities. Our results show that can many by sample training from knowledge-base.