Content area
Full Text
Search engines are among the most successful applications on the web today. So many search engines have been created that it is difficult for users to know where they are, how to use them, and what topics they best address. Metasearch engines reduce the user burden by dispatching queries to multiple search engines in parallel. The SAVVYSEARCH metasearch engine is designed to efficiently query other search engines by carefully selecting those search engines likely to return useful results and responding to fluctuating load demands on the web. SAVVYSEARCH learns to identify which search engines are most appropriate for particular queries, reasons about resource demands, and represents an iterative parallel search strategy as a simple plan.
Companies, institutions, and individuals must have a presence on the web; each are vying for the attention of millions of people. Not too surprisingly then, the most successful applications on the web to date are search engines: tools that assist users in finding information on specific topics.
A variety of search engines are available, from general, robot based (for example, ALTAVISTA [www.altavista.digital.com] and WEBCRAWLER [webcrawler.com]) to topic or area specific (for example, FTPSEARCH [ftpsearch. unit.no/ftpsearch] and DEJANEWS [WWW. dejanews.com]). Each uses different algorithms for collecting, indexing, and searching links; thus, each returns different results for similar queries. Empirical results indicate that no single search engine is likely to return more than 45 percent of the relevant results (Selberg and Etzioni 1995). To find what they desire, users might need to query several search engines; metasearch engines automate this process by simultaneously submitting a single query to multiple search engines.
The simplest metasearch engines are forms that allow the user to indicate which search engines should be contacted (for example, ALL-IN-ONE [www.albany.net/allinone] and METASEARCH [members.gnn.com/infinet/ meta.htm]). PROFUSION (Gauch, Wang, and Gomez 1996; www.designlab.ukans.edu/profusion) gives the user the choice of selecting search engines themselves or letting PROFUSION select three of six robot-based search engines using hand-built rules. METACRAWLER (Selberg and Etzioni 1995; metacrawler.cs.washington.edu:8080/home.html) significantly enhances the output by downloading and analyzing the links returned by the search engines to prune out unavailable and irrelevant links.
Metasearch engines reduce the burden on the user. They make available search engines that might have been unknown to the user. They handle the simultaneous submission...