The role of ontologies in autonomic computing systems (19 Sep 2004)
Within the scope of IBM's research and development activities, several correlation engines have been developed and successfully applied in various domains. However, the increasing importance of such self-managed systems in an intra-enterprise environment requires their deep semantic interoperability. Moreover, such integration is of primary importance for on demand computing, one of IBM's strategic goals.
In order to achieve this semantic integration, we have proposed in this paper the usage of ontologies as the conceptual backbones of correlation engines. We see several benefits of using such an approach: First, in the area of domain modeling ontologies facilitate interoperability between correlation engines by providing shared understanding of a problem domain. Second, ontologies provide the formalization of shared understanding necessary to make such understanding machine-processable. Such machine processability is the basis for the next generation of the World Wide Web, the so-called Semantic Web, allowing us to achieve compliance with existing Web standards. Finally, the explicit representation of the semantics of data, in combination with ontologies, enables correlation engines to provide a qualitatively new level of services in autonomic computing systems.
In order to illustrate these benefits, we have presented a case study for the eAutomation correlation engine. We showed how an ontology-based correlation engine can extend the capabilities of the original engine by exploiting the properties of an ontology, namely better modeling of the underlying domain, especially with respect to the expression of axiomatic knowledge, higher reusability, easier extensibility of the system, and the possibility of formal verification of the system.
Finally, we showed how the use of ontologies can help to improve the effectiveness of autonomic systems. In particular, ontologies enable additional levels of services such as justification, ranking of proposed solutions to a failure, and gap analysis.
Article URL: http://www.research.ibm.com/journal/sj/433/stojanovic.html
Read 201 more articles from IBM sorted by
Next Article: Towards the next generation of enterprise search technology