Time Adaptive Boosting Model for Topic Tracking
Huizhen Wang, Jingbo Zhu, Duo Ji, Na Ye, Bin Zhang · 2006
The technology of topic tracking can help people find what they are interested from the vast information sea. Since topics develop dynamically, topic excursion problem may appear in the tracking process. To overcome this problem and the shortcomings of current adaptive methods, we propose a new adaptive method for topic tracking. We call it time adaptive boosting (TAB) model. This model adopts the idea of boosting and presents new algorithm to the adaptive learning mechanism in the task of topic tracking. This algorithm can solve the problem of topic excursion, and remedy the deficiency of current adaptive methods. Time sequence of topic tracking task is also considered in the algorithm. We use sigmoid function to express it. In experiments we use the Chinese part in TDT4 corpus as test corpus, and use the TDT2004 evaluation metric to evaluate the adaptive Chinese topic tracking system based on TAB. Experimental results show that the adaptive method based on TAB can improve the performance of topic tracking.