An expert system hybrid architecture to support experiment management

•We designed an architecture for expert systems supporting experiment management.•The architecture integrates declarative, procedural and workflow oriented approaches.•We developed a proper ontology for knowledge and workflow organization.•We presented a bioinformatics scenario that exploits the architecture’s main features.

Specific expert systems are used for supporting, speeding-up and adding precision to in silico experimentation in many domains. In particular, many experimentalists exhibit a growing interest in workflow management systems for making a pipeline of experiments. Unfortunately, these type of systems does not integrate a systematic approach or a support component for the workflow composition/reuse. For this reason, in this paper we propose a knowledge-based hybrid architecture for designing expert systems that are able to support experiment management. This architecture defines a reference cognitive space and a proper ontology that describe the state of a problem by means of three different perspectives at the same time: procedural, declarative and workflow-oriented. In addition, we introduce an instance of our architecture, in order to demonstrate the features of the proposed work. In particular, we model a bioinformatics case study, according to the proposed hybrid architecture guidelines, in order to explain how to design and integrate required knowledge into an interactive system for composition and running of scientific workflows.