Anzeige
Erschienen in:
2016 | OriginalPaper | Buchkapitel
Detecting Live Events by Mining Textual and Spatial-Temporal Features from Microblogs
verfasst von : Zhejun Zheng, Beihong Jin, Yanling Cui, Qiang Ji
Erschienen in: Web-Age Information Management
Aktivieren Sie unsere intelligente Suche, um passende Fachinhalte oder Patente zu finden.
Wählen Sie Textabschnitte aus um mit Künstlicher Intelligenz passenden Patente zu finden. powered by
Markieren Sie Textabschnitte, um KI-gestützt weitere passende Inhalte zu finden. powered by
Abstract
As microblogging services on the mobile devices are widely used, microblogs can be viewed as a kind of event sensor to perceive the dynamic behaviors in the city. In particular, detecting live events in microblogs, such as mass gathering, emergencies, etc., can help to understand what happened from the point of view of people who are present. For identifying the live events from a large number of short and noisy microblogs, the paper builds a generative probabilistic model named the ST-LDA model to cluster the microblogs whose semantics, time and space are similar into the same topic, and then determines the live events from the topics by an HMM-based method. The paper conducts the experiments on the real microblogs from weibo.com. Experimental results show that our method can detect live events more accurately and more completely than the LDA-based method and the TimeLDA-based method.