Enterprise Architecture Model Analysis Using Description Logics

Enterprise architecture involves the use of models for facilitating the communication between stakeholders, which is required for the maintenance of the business/IT alignment. In that sense, the models should provide the necessary information for articulated decision-making, which, in complex scenarios, can result in the overwhelming complexity of the architecture descriptions. The complexity of the models might render manual analysis practically impossible, hence the appearance of automated analysis proposals throughout the years. Ontologies are a means for formally describing domain models, adding unambiguous meaning to its terms and relations. Moreover, ontologies provide computational inference mechanisms which can be used for the analysis of domain models. This work discusses the use of ontologies for EA analysis, focusing on the particular case of description logics. For that purpose, a survey on analysis approaches is performed and the reasoning features provided by description logics are matched to the different types of EA analysis. Based on such findings, a set of constructs is proposed for the representation, integration, and analysis of EA models. Finally, a demonstration of such constructs applied in compliance analysis is performed using an example scenario.