Medical Ontologies as a Knowledge Repository

摘要

We analyzed existing medical ontologies from a view point of academic domain ontology research and suggest future directions of medical knowledge repository systems. Academic domain ontology is a form of knowledge repository system which enables both human and machine agents to store, systematize and utilize the most complex accumulation of knowledge resources. From our previous attempts to construct academic domain ontolotgies, we focus on the following three aspects of knowledge repository and apply them to an analysis of medical ontologies. (1) Local nature of knowledge whose range extends from private level to public level is an essential feature of academic knowledge. Proper knowledge repository system needs to support the feature. (2) Collective acquisition of knowledge could help the knowledge repository construction in great degree. Human experts could participate in a collection of knowledge with a distributed input manner. (3) Usable knowledge repository needs to provide relational knowledge between pieces of knowledge (communality and differentiation between them), and meta knowledge about a piece of knowledge (how to use it). Citing examples from medical ontologies such as UMLS, SNOMED, GALEN and MEDIS, we argue the principles for academic medical domain ontology for medical communications.