ABSTRACT

This paper explores the use of hierarchical structure to classify a heterogeneous collection of web pages. In the hierarchical classification, a model learns to distinguish a second level category from all other categories that are within the same top level. In the flat non hierarchical classification, a model distinguishes a second level category from all existing second level categories. We use Naïve Bayes classifier which has been proved to be effective for web content classification, but has not been previously explored in the case of hierarchical classification. This paper analyses the feasibility of a web page classifier which exploits the hierarchical structure of categories and studies their recall, precision and F-measure scores.