Ontology Engineering: An Application Perspective

Enterprise Integration is a key challenge faced by many organizations, including those of the U.S. Government. An important enabler of enterprise integration is data integration. There is promise in applying Semantic Web technologies to enable data integration of disparate data sources based on the semantics (meaning) of the data. As we start to build ontologies within United States (U.S.) government domains for this purpose, we consider how Semantic Web technologies may operate within existing operational infrastructures and how best to build ontologies to facilitate usability. This paper explores the ontology engineering and adoption challenges based on current state of the art in Semantic Web technology. We consider alternative approaches to engineer ontologies, discuss current and emerging standards in this area, look at approaches to integrate data through ontology mapping, and outline a set of skills necessary to develop and apply ontologies. We then propose an architecture to insert Semantic Web technologies into an existing infrastructure to enable data and application interoperability, based on experimentation within the Intelligence Community. The paper concludes with recommendations on a way ahead, and highlights some potential pitfalls for future ontology developers.