Civil Transportation Event Extraction from Chinese Microblog

People produce hundreds of millions of microblogs everyday. With its 140-character message, Microblog has yielded an enormous corpus of information, which is noisy but informative in some way. However, previous work with standard NLP tools of event extraction performs poorly on Microblog. In this paper, we adopt a series of methods to extract events from Chinese microblogs. In particular, we grab the chatters from Sina Weibo to extract civil transportation information. We eliminate buzz in Weibo, use CRF methods to filter microblogs so as to focus on transportation, and we also use CRF to recognize named entities and to extract events.