Related Document Discovery

Search engines that access pre-indexed documents are able to quickly refer users to related documents. However, the problem becomes harder when the search engine is federating sources in real-time with no access to pre-indexed data. This project will design, build, and deploy a federated search application that quickly analyzes summary information from documents returned from a search and, in real-time, generates and executes alternative queries that return related documents. Commercial Applications and other Benefits as described by the awardee: A related document discovery capability should compete successfully with crawl-based search engines that already provide this capability. The serious researcher would benefit from the ability to view documents that his/her original search may not have identified as relevant.