Concept similarity and cosine similarity result merging approaches in metasearch engine

Abstract

Metasearch engines provide a uniform query interface for Internet users to search for information. Depending on users need, they select relevant sources and map user queries into the target search engines, subsequently merging the results. In this paper, we have proposed a metasearch engine, which have two unique steps (1) searching through surface and deep web, and (2) Ranking the results through the designed ranking algorithm. Initially, the query given by the user is given to the surface and deep search engines. Here, the surface search engines like Google, Bing and Yahoo are considered. At the same time, the deep search engine such as, Infomine, Incywincy and CompletePlanet are considered. The proposed method will use two distinct algorithms for ranking the search results, which are concept similarity and cosine similarity. © 2013 Springer Science+Business Media New York.