Automated Classification of Web Sites using Naive Bayesian Algorithm
Ajay S. Patil, B. V. Pawar · 2012
Abstract— Subject based web directories like Open Directory Project's (ODP) Directory Mozilla (DMOZ), Yahoo etc., consists of web pages classified into various categories. The proper classification has made these directories popular among the web users. The exponential growth of the web has made it difficult to manage human edited subject based web directories. The World Wide Web (WWW) lacks a comprehensive web site directory. Web site classification using machine learning techniques is therefore an emerging possibility to automatically maintain directory services for the web. Home page of a web site is a distinguished page and it acts as an entry point by providing links to the rest of the web site. The information contained in the title, meta keyword, description and in the labels of the anchor (A HREF) tags along with the other content is a very rich source of features required for classification. Compared to the other pages of the website, webmasters take more care to design the homepage and its content to give it an aesthetic look and at the same time attempt to precisely summarize the organization to which the site belongs. This expression power of the home page of a website can be exploited to identify the nature of the organization. In this paper we attempt to classify web sites based on the content of their home pages using the Naive Bayesian machine learning algorithm.