KACTL: Knowware based automated construction of a treelike library from web documents

Abstract

This paper proposed a knowware based supervised machine learning technique for domain specific regression and classification of Web documents. It is simple because it is only based on word counting techniques without natural language understanding and complicated statistic techniques. Starting from constructing a domain sub-division tree and assigning a training set of documents to its nodes, the algorithm produces a labeled classification tree with a characteristic vector for each node. This tree is used to classify any number of documents in that particular domain. A tool for developing Web portal is also provided to build a Web station for displaying the final treelike library of documents. © Springer-Verlag Berlin Heidelberg 2012.