Publications
Expert Knowledge Base Designed Using ER-Modeling Technique
10/01/1986
Expert Knowledge Base Designed Using ER-Modeling Technique
Frank Naeymi-Rad1, Tim Koschmann4, David Trace1, Theresa Kepic3,
C. Robert Carlson2, Max Harry Weil1, Martha Evens2
1University of Health Sciences/The Chicago Medical School, North Chicago IL 60064
2Illinois Institute of Technology, Department of Computer Science, Chicago IL 60616
3Department of Obstetrics/Gynecology, Cook County Hospital, Chicago, IL 60612
4Xerox AI Systems, Arlington Heights, IL 60005 U.S.A.
Abstract
A DBMS modeling technique, known as the ER (Entity Relationship) - Diagram, was modified to design a database for treatment protocols to be used in conjunction with a Knowledge Base for a computer based diagnostic consultant. The diagnostic consultant, MEDAS, is a pattern recognition system designed for use in an emergency room setting. The technique of ER diagramming was selected because of its ease of use and domain independence. Moreover, ER-Diagramming was demonstrated to apply equally well to different AI knowledge representations. A recommendation is made for an updated model to support functional mapping and relational network with the same diagram.
Download this publication as a PDF file