Automated Online News Classification with Personalization

Classification of online news, in the past, has often been done manually. In our proposed Categorizor system, we have experimented an automated approach to classify online news using the Support Vector Machine (SVM). SVM has been shown to deliver good classification results when ample training documents are given. In our research, we have applied SVM to personalized classification of online news. In personalized classification, users can define their personalized categories using a few keywords. By constructing search queries using these keywords, Categorizor obtains both positive and negative training documents required for the construction of personalized classifiers. In this paper, we describe the preliminary version of Categorizor and present its system architecture.