A Multi-Formalism Framework To Generate Diagnostic Decision Support Systems

The task of a Diagnostic Decision Support System (DDSS) is to deduce the health status of a physical system. In this paper, a multi-formalism framework to generate DDSS software based on formal descriptions of the application domain and the diagnostic computations is proposed. 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 dual-formalism descriptions, while the structure of the DDSS is invariant across applications. An evaluation involving an artificial scalable domain related to the diagnosis of air conditioning systems is presented to exemplify and to test the proposed framework.