Study on Key Technology of Topic Tracking Based on SVM

Abstract

Text classification is the key technology for topic tracking, and vector space model (VSM) is one of the most simple and effective model for topics representation. On the basis of VSM and support vector machines (SVM), we have studied how feature space dimension in VSM as well as linearly separable and non-separable SVM affect topic tracking. Then we get the variation law that they affect topic tracking, and add up their optimal values in topic tracking. Finally, TDT evaluation method proves that optimal topic tracking performance based on linearly separable SVM increases by 4.522% more than linearly non-separable SVM.