KeyConcept (Conceptual Search Engine)

University of Kansas
posted on 10/20/2005

This search engine retrieves documents based upon a combination of keywords and conceptual matching. Documents are automatically classified to determine the concepts to which they belong. Query concepts are determined automatically from a small description of the query or explicitly entered by the user. Concepts can be determined by means of a user profile; we have built a system that creates user profiles based on implicitly collected information, specifically the queries submitted and snippets of user-selected results.

Suggested Uses

Added to current search engines, KeyConcept will increase their effectiveness as users will receive search results based on their interests, not web site popularity.

Detailed Description

This search engine retrieves documents based upon a combination of keywords and conceptual matching. Documents are automatically classified to determine the concepts to which they belong. Query concepts are determined automatically from a small description of the query or explicitly entered by the user. Concepts can be determined by means of a user profile; we have built a system that creates user profiles based on implicitly collected information, specifically the queries submitted and snippets of user-selected results. This system provides search results more relevant to the user's current activities and tasks.

Limitations

None known.

File Number: NCL-0106

Web site: http://www.ittc.ku.edu/keyconcept/

Other Information: *Limitations* None known. *Third Party Rights* None. *State of Development* Development is complete. Technology is ready for licensing. A demo is available at http://www.ittc.ku.edu/keyconcept/. *Testing* Completed.