The web is expanding exponentially. In January 2007, there were nearly 30 million pages (WWW FAQ, 2007). This expansion has led to reliance on search engines to find web resources. This in turn casts responsibility on the search engines to meet the needs and expectations of the scholarly community. Using more than one search engine is futile if overlapping is frequent and substantial. Overlapping is genuine if the common results are highly relevant to the user's query. Use of different search engines simultaneously reduces searching time and increases efficiency. Though search engines index multiple and separate resources, some results occur in many search engine's databases and in some cases a search engine retrieves results by indexing other search engines' databases. The present study is an attempt to identify search engines with less overlapping for use by the scholarly community.
In the ocean of literature on search engines features, precision, recall, and other technical aspects, there has been little attempt to study overlap. Bharat and Border (1998) measured overlap among websites indexed by Hotbot, Altavista, Excite, and Infoseek using 10,000 queries carried out at two different intervals of time in June 1999 and November 1999, and found that the overlap was very small, less than 1.4 percent of the total coverage. Ding and Marchionini (1998) evaluated results retrieved by Infoseek, Lycos, and Opentext to measure the level of common results and report a low level of overlap. Chignell, Gwizdka, and Bonder (1999) found little overlap in the results returned by various search engines and describe meta-search engines as useful. Gordan and Pathak (1999) studied five search engines by measuring overlap at a document cutoff value of 20, 50, 100, and 200 and find that approximately 93 percent of the results were retrieved by only one search engine. Nicholson (2000) replicated the 1998 Ding and Marchionini study and found similar results with low web search engine overlap. Ferrara, da Silva, and Delgado (2004) evaluated previous overlap studies with the finding that documents retrieved by multiple information retrieval systems in relation to the same query are more likely to be relevant. Spink, Jansen, Kathuria, and Koshman (2006) examined the overlap among results retrieved by three major web search engines (Google, Ask Jeeves, and Yahoo) using a set of 10,316 randomly selected queries. The study shows that the...