| The internet is the main and most reliable source for any individual to be updated with evolving technology. The big question here is that whether people are able to locate the exact information they were looking for? Most people search the internet with the help of certain well known search engines, most of which are commercial, to find the results. However, the current search techniques lead the user to a result based on the popularity of the page or based on how reliable the content is. These search results are the results that the search engine estimates to be the correct ones. The web pages retrieved with such techniques might give the user certain good results. But in most cases they are not the results for which the user was looking. Therefore the user tends to look at the world from the search engine's perspective. This situation can be enhanced by the help of faceted search. The crux in implementing faceted search involves processes like categorization, implementation of passive and active filters, etc. In this thesis an algorithm has been developed to implement categorization, which is the fundamental step in implementing faceted search. The Naive Bayesian classifier is used to classify the web content into categories, based on which the faceted search could be implemented. To validate the algorithm, it is applied on a small training data comprising of URLs with known content. The classification results were perfect with 100 percent accuracy. |
