Ontologies provide a domain-specific approach to formally name and define system types, properties, and entity interrelationships. Outside traditional semantic web ontological use, this paper shows how ontologies can be used to define and capture the processes involved in information systems development. These can then be analyzed and used for future requirements engineering. The authors introduce the i* ontology model and show its use as a process language and ontology retrieval capability. This i* ontology defines an extensive set of binary relationships that operate on tasks, actors, goals, and resources. This is called the i* strategic dependency (SD) model, implemented in version 2 of the Web Ontology Language (OWL2), and consists of 28 concepts, 57 relations of which 26 are taxonomic (that is, “IS-A”) relationships, and five attributes.
The authors introduce a three-stage approach for analyzing the value and extracting a chosen ontology in the information systems development domain. Stage one defines the theoretical mapping where early requirement models are converted into an intermediate retrieval ontology. Together with the SD model, a strategic relational (SR) i* model illustrates the required domain axioms. These actor-required relationships, such as actor-to-goal or actor-to-task, are then encoded in the OWL ontology for use in either stage two or stage three.
Stage two performs consistency checks where the rigor depends on the level of formal modeling. An OWL2 reasoner, HermiT reasoning engine, is used to ensure consistency, for example, between the individuals and their descriptions. Stage three uses the i* ontology model to identify and retrieve a candidate set of relevant end-use ontologies.
A more extensive example of a meeting scheduler is encoded in the i* retrieval ontology. All three stages are described in the paper in some detail, showing how this approach can be an improvement when defining requirements or checking consistency.