Next: The Cheshire II Up: The Cheshire II Previous: Merging Ranked Results

Browsing and Relevance Feedback

The Cheshire II system facilitates several methods of open-ended, exploratory browsing through following dynamically established linkages between records in the database, in order to retrieve materials similar to those already found. The search engine supports probabilistic retrieval based on any items that have already been seen and selected by a user. Thus, any citation or document seen by the user can become the basis for a ``nearest neighbor'' search, where the contents of the selected record are fed back to the probabilistic search algorithm to find those records in the database most similar in content to the one specified. The user's selection of relevant citations can also be used to refine the initial search statement and automatically construct new search statements for relevance feedback searching.

In addition, the graphic user interface supports dynamically generated hypertext linkages from elements in a given record (such as subject headings or call numbers in a display) to other records in the source database though automatic Boolean query formulation.

Contact: Ray R. Larson