The implementation of software systems is often based on a design captured by business processes. The advantage of this is that a domain expert can easily understand the operation of the system. In such systems, some important question can be raised, for example: How to prove that the business process meets the specifications? How to prove that the operation of the system meets the specifications? How to evaluate the system in terms of IT and business performance and dependability? How to detect the failures of the serving systems?
Analysis of system logs can help answer these questions. After many run or long operational time, huge amount of data can be produced. In this case, the analysis may become difficult therefore an effective tool is needed, which supports the experts' work.
To solve the problem of handling huge amount of data, we created a diagnostic environment. In my thesis I present an environment, which is based on an industrial project. In this environment the business expert has the possibility to perform analysis by writing rules in a domain specific language, generated from the process model. These will be evaluated on events, generated by previously executed process.
The rules should be checked since the successful diagnostic depends on the rules' quality. In the thesis I demonstrate the correctness and completeness of the rulebase through analysis. For instance, if an expert notes some critical parts of a model, then it can be checked whether the business analytic has written rules for all important cases. In the design of diagnostic rulebase it is important to check the consistency of the rules since usage of controversial rules can lead to an incorrect, false statement. The semantic analysis of the rules is essential to make correct diagnostics.
Another way to ensure the rules' quality to generate them. Many simple rule can be generated, which also can help analysts' work.