The role of ontologies in autonomic computing systems

The goal of IBM's autonomic computing strategy is to deliver information technology environments with improved self-management capabilities, such as self-healing, self-protection, self-optimization, and self-configuration. Data correlation and inference technologies can be used as core components to build autonomic computing systems. They can also be used to perform automated and continuous analysis of enterprise-wide event data based upon user-defined configurable rules, such as those intended for detecting threats or system failures. Furthermore, they may trigger corrective actions for protecting or healing the system. In this paper, we discuss the use of ontologies as a high-level, expressive, conceptual modeling approach for describing the knowledge upon which the processing of a correlation engine is based. The introduction of explicit models of state-based information technology resources into the correlation technology approach allows the construction of autonomic computing systems that are capable of dealing with policy-based goals on a higher abstraction level. We demonstrate some of the benefits of this approach by applying it to a particular IBM implementation, the eAutomation correlation engine.