User-based meta-search with the co-citation graph

Although there are numerous search engines in the Web environment, no one could claim producing reliable results in all conditions. This problem is becoming more serious considering the exponential growth of the number of Web resources. In the response to these challenges, the meta-search engines are introduced to enhance the search process by devoting some outstanding search engines as their information resources. In recent years, some approaches are proposed to handle the result combination problem which is the fundamental problem in the meta-search environment. In this paper, a new merging/re-ranking method is introduced which uses the characteristics of the Web co-citation graph that is constructed from search engines and returned lists. The information extracted from the co-citation graph, is combined and enriched by the userspsila click-through data as their implicit feedback in an adaptive framework. Experimental results show a noticeable improvement against the basic method as well as some well-known meta-search engines.