A Dynamic Event Tracking Model Combining With Time Factor

The information associated with an event is scattered in different places and differenttimes, and the event tracking technology can organize the information effectively on thenetwork to help people to fully grasp the development of the event. However, in eventtracking, the related stories constituting the initial event term set are few and the content ofthe relevant documents evolves dynamically, which leads to the inaccurate the event term set.Concerning the problem in the event tracking, the study put forward a dynamic modelfor event tracking with reference to the time information. The dynamic model introduced thetime factor into the traditional vector model to get the time similarity of the samecharacteristic words between the document and the event, and then applied the time similarityto calculate the similarity of the document and the event. If a document was related to theevent, the new characteristic words in the document would be re-ajusted. The experiment wasevaluated by Detection Error Tradeoff(DET), and the results show that the dynamic model forevent tracking improves the system performance effectively, and its minimum normalizedcost of tracking loss is reduced by about9%.