Relevance feedback is a technique used in computer search systems whereby the user is asked to rank or otherwise indicate which of the documents returned from an initial search best match his or her needs. The search system then attempts to refine the search based on those identified documents.
There are many methods, techniques, and algorithms by which the system might analyze the documents and "plow the data" back into a search. One method is to use term weighting techniques--explained elsewhere in this guide.
Return to the main menu.