A meta search approach to find similarity between web pages using different similarity measures

Abstract

Search engines are the online services available, which are used to locate necessary information on World Wide Web. As the web is growing at a very rapid rate, the pages that are similar to each other are also increasing. Hence, it is better to have a system that can discover similar web pages. In this paper, A Meta search approach is applied for the information retrieval purpose which retrieves pages from the result list of different search engines and content present in the web pages is analyzed on the basis of which system finds similarity between them. Web pages are represented in vector space which represents each web document as a vector and the terms present in that webpage as its components. Similarity is computed by using different similarity measures i.e. Cosine Similarity, Jaccards Coefficient and Dice Coefficient. A comparative analysis of these similarity measures is done to find out which measure performs better in terms of precision as well as recall. © 2011 Springer-Verlag Berlin Heidelberg.