Naive Bayes Approach for Website Classification

World Wide Web has become the largest repository of information because of its connectivity and scalability. With the increase in number of web users and the websites, the need for website classification gains attraction. The website classification based on URLs alone plays an important role, since the contents of web pages need not be fetched for classification. In this paper, a soft computing approach is proposed for classification of websites based on features extracted from URLs alone. The Open Directory Project dataset was considered and the proposed system classified the websites into various categories using Naive Bayes approach. The performance of the system was evaluated and Precision, Recall and F-measure values of 0.7, 0.88 and 0.76 were achieved by this approach.