The New Jersey Institute of Technology's
Electronic Theses & Dissertations Project

Title:	Object oriented partitioning for a medical vocabulary based on semantic network
Author:	Singh, Mansnimar
View Online:	njit-etd1997-091 (x, 98 pages ~ 3.4 MB pdf)
Department:	Department of Computer and Information Science
Degree:	Master of Science
Program:	Computer Science
Document Type:	Thesis
Advisory Committee:	Perl, Yehoshua (Committee chair) Geller, James (Committee member) Halper, Michael (Committee member)
Date:	1997-05
Keywords:	Natural language processing (Computer science). Medicine--Information services.
Availability:	Unrestricted
Abstract:	Computers have become ubiquitous and indispensable part of everyday life both for personal use and in the workplace. Medicine is one of the few domains that has not fully adopted computerization. One impediment to this emanates from a communication gap between the computer science professional and the medical professional. Besides, medical terminology is full of synonyms and medical professionals use them according to their personal preferences. This lack of common terminology has prevented sharing of knowledge and automating data processing, resulting in the healthcare information explosion. Semantic network models have been developed to represent medical concepts and to provide a common repository of medical terms. These networks are huge collections of terms and deal with the concepts individually. The semantic models, however, have not resolved management and comprehension difficulties associated with large number of terms and their complex semantics. Object-Oriented modeling, as demonstrated in this thesis, provides a mature technology to model complex concepts for a computerized medical vocabulary. Object class abstraction, that represents objects in the same context, promises to solve the comprehension and management problems. Such a vocabulary would facilitate exchange of information, data processing and building decision support systems.