Software Engineering (SE) is a wide domain, where ontologies are useful instruments for dealing with semantics and knowledge management related problems. When SE ontologies are built and used in isolation, some problems remain, in particular those related to knowledge integration. Over the years, we have produced several ontologies related to SE subdomains, such as Software Measurement, Software Process, Requirements, Configuration Management, Software Project Management, and Software Test. We experienced some problems, such as the same concept appearing with different meanings in different ontologies and the same term being used to designate different concepts in different ontologies, among others.
In large and complex domains, as it is the case of SE, if we try to represent the whole domain as a single ontology, we will achieve a large and monolithic ontology that is hard to manipulate, use, and maintain. On the other hand, representing each subdomain separately would be too costly, fragmented, and again hard to handle. Therefore, we have investigated how to develop and organize ontologies in an architecture that enables integrating existing ontologies and adding new ontologies, keeping consistency between them.
We propose to organize SE ontologies in an ontology network that supports creating, integrating and evolving its ontologies. An ontology network is a collection of ontologies related together through a variety of relationships, such as alignment and dependency. A networked ontology, in turn, is an ontology included in such a network, sharing concepts and relations with other ontologies.
SEON is a Software Engineering Ontology Network, which provides a well-grounded network of SE reference ontologies, and mechanisms for building and integrating new subdomain ontologies into the network.
SEON architecture is organized in layers. Briefly, in the background, we have UFO (Unified Foundational Ontology) to provide the general ground knowledge for classifying concepts and relations in the network. In the center, core ontologies are used to represent the general domain knowledge, being the basis for the subdomain networked ontologies. Finally, going to the borders, (sub)domain ontologies appear, describing more specific knowledge.
Being an ontology network, SEON is like a living organism and is constantly evolving. It requires a continuous and long-term effort with ontologies being added and integrated incrementally.
SEON was introduced in this paper:
SEON specification is available at SEON website. The specification presents all networked ontologies, the network graphic and also a machine-processable lightweight version implemented in OWL.