Comprehending the Structure of a Medical Vocabulary Using Object-oriented Database Modeling

The purpose of this paper is to demonstrate how the transformation of a medical vocabulary based on a Semantic Network (SN) model into a vocabulary based on an Object-Oriented Database (OODB) model helps in the maintenance of the vocabulary. We describe an OODB schema which captures the overall structure of the vocabulary in a compact form and uncovers some errors and inconsistencies made in the vocabulary's original modeling. The resolution of these mistakes leads to an improved version of the SN-based vocabulary. A new OODB schema for the vocabulary is then derived based on the improved SN version. This experience demonstrates how the abstraction and modeling capabilities of OODBs can be used to enhance a user's understanding of the overarching structure of a complex medical vocabulary system. The OODB schema developed herein serves as the basis for the Object-Oriented Healthcare Vocabulary Repository (OOHVR), a medical vocabulary implemented as an ONTOS database.