Document Retrieval Using Fuzzy Matching and Aggregation

Ronald R. Yager

A basic idea inherent in the use of fuzzy sets technology is the concept of computing with words. This ability will enable developers of future information retrieval systems to provide more intelligent and human friendly systems by allowing uses to interact with these systems in natural language while at the same time providing a machinary which will allow the computer to perform the types of formal operations needed. The central component of any document retrieval system is a library of documents which can be text, audio, image, video or any combination. In order to retrieve documents from this library some matching mechanism is required. To facilitate this matching process we must introduce a collection of features to help distinguish the objects in the library. Using these features each document in the library can be identified by an index consisting of values of the features associated with the document. A search is initiated by introducing a probe consisting of values for the features describing the users interest. The feature values in the probe are matched with the corresponding features of the objects in the library. Those documents which score the highest matching with the probe are then selected. In attempting to describe features associated with a particular environment considerable use can be made of the fuzzy set technology in that it provides a formal mechanism to enable us to more naturally represent the imprecision and gradedness associated with the concepts used in feature description.

This page is copyrighted by AAAI. All rights reserved. Your use of this site constitutes acceptance of all of AAAI's terms and conditions and privacy policy.