A Multilayer Ontology to represent Business Process Models and Execution Data

Abstract

Business process management is still characterized by a low degree of mechanization. One of the main causes is a lack of an appropriate semantic representation for business process models and execution data unified in a single ontology. Many researchers addressed this problem but the ontologies proposed so far define a proprietary property called occurrenceOf" to link a process model to its execution data. In this paper I propose an alternative approach based on a multilayer representation using the instance Of relation instead. This is cognitively more adequate and allows to use standard reasoning mechanisms. I show how OWL-FA enables a semantic representation of the metamodel in an ontology, providing decidable reasoners. A case study is used to clarify some representation problems and the semantic issue is addressed through a detailed analysis. The results section shows how it is possible to answer some queries with the support of this ontology, demonstrating how our ontology can be a suitable tool to enhance the mechanization of business process management. The last part of this thesis is dedicated to possible extensions through integration with other ontologies in order to enable the user to achieve a more complete process definition using other modeling languages. In this case I use Archimate through ArchiMEO, an RDFS ontology developed by FHNW, to represent some fundamental concepts of the case study.