Effective derivation of test cases from requirements

Software quality assurance is meant to achieve required quality in product and process. The most used method for analytical quality assurance in industrial settings is testing. Testing is the process of judging on the correctness of a system by feeding it some input test cases and observing its reactions. Executing tests on software facilitates to detect the presence of faults, thus providing a practical means for assessing software correctness. The development of test cases based on the system-level requirements is a key part of system and user acceptance testing. Scenario-based testing is a paradigm shift in this field. Instead of examining individual functions in isolation, scenario-based testing focuses on verifying the system behavior within the scope of entire usage scenarios from the viewpoint of the actual user. The currently known approaches for generating system test cases based on scenarios suffer from several flaws. Thus, despite widely acknowledged benefits of system-level testing based on system requirements, current software testing practices mainly focus on generating test cases from the actual code of the system. Aiming to put functional system and acceptance testing on a more rational and systematic basis, the goal of this thesis is to provide test engineers a comprehensive method for deriving functional test cases from system-level requirements depicted in terms of use cases. An integrated and comprehensive approach and supplemental guidelines for efficiently developing effective test cases is presented. Beyond development of test cases the proposed approach permits to systematically analyze use cases in detail in order to discover ambiguities, errors, and omissions resulting in a high quality requirements document.