A Comprehensive Study of Features and Algorithms for URL-Based Topic Classification
ACM Transactions on the Web (TWEB)
Hi-index | 0.00 |
Explosive growth of the Internet makes it difficult for search engines to give relevant results to the users within a stipulated time. Search engines store the web pages in classified directories and for this process even though some search engines depend on human expertise; most of the search engines use automated methods for classification of web pages. In this paper we use machine-learning techniques for the automated classification of web pages. We consider only URL features for classification as the URL name is unique, meaningful and helps identification of their subject categories most of the times. Experimental results show that machine learning techniques for automated classification of web pages with URL features proves to be the best and more useful method for search engines.