Samson Tu,
Senior Research Scientist,
Staff Emeritus Retiree,
Stanford University

Abstract:
The boundaries between ontologies, information models, and knowledge bases are often blurred and people are often confused about the representational choices that are appropriate for each. Over the years, numerous formalisms¾semantic networks, rules, frames, UML, RDF, formal logic¾have been used to make computational artifacts. In this talk, I will use examples in health informatics to illustrate some critical distinctions. Some formalisms provide templates for modeling; others formulate assertions as axioms. Some make the open-world assumption while others assume a closed world. Such distinctions may become pitfalls when formalisms and requirements are mismatched. I will argue that the current suite of tools do not provide the integration that we need to develop intelligent applications and discuss some emerging technologies that may ameliorate the situation.