Independent testing teams use requirements as the basis to develop test cases and automated test scripts. The projects are executed under severe schedule constraints, due to which, the testers have to focus their testing efforts on error-prone and important features. Numerous source code based techniques for identifying error-prone features/components have been developed. However, they are based on source code analysis. Independent testing teams rarely have access to source code and they find

more »

... difficult to use code based techniques. In many cases, the domain experts use Business Process Model and Notation (BPMN) to represent the business requirements. In this paper, we propose an approach to identify error-prone transactions in enterprise applications using a BPMN. It helps in distinguishing between source code errors and test script errors. We have adapted this approach from an existing source code based technique. Our experiments with the approach show that it can identify the location of actual as well as seeded errors in both source code and test scripts.