Towards an ontology-based framework to generate Diagnostic Decision Support Systems

Abstract

The task of a Diagnostic Decision Support System (DDSS) is to deduce the health status of a physical system. We propose a framework to generate DDSS software based on formal descriptions of the application domain and the diagnostic rules. The key idea is to describe systems and related data with a domain ontology, and to describe diagnostic computations with an actor-based model. Implementation-specific code is automatically generated from such descriptions, while the structure of the DDSS is invariant across applications. Considering an artificial scalable domain related to the diagnosis of air conditioning systems, we present a preliminary experimental comparison between a hand-made DDSS, and one generated with a prototype of our framework. © Springer International Publishing Switzerland 2013.