Improving Trace Analysis Using Ontologies for Hardware Resourcing

Abstract

Testing is one of the traditional techniques used to verify the quality of complex systems. Traditionally, black-box testing relies on the degree of controllability and observability of the system under test; a system with increased controllability and observability is a system easier to test. One common observation point for testing is execution traces. Execution traces are sequences of events representing observation of the system under test. The execution traces are usually stored as plain text files (i.e., logs). The current size and complexity of systems makes the execution trace analysis a complex and time consuming task given the size and format of the information. This paper presents the application of ontological methods in facilitating execution trace analysis by defining an initial Execution Trace Ontology that is used by different ontology query tools. The queries used over the ontology allow us to identify errors presented in the execution trace associated with different aspects of the case study. Results showed the feasibility of this approach, where ontologies helped to provide semantics to the information and reasoning engines (ontology query engines), facilitating the definition of test goals.