Automatic test case generation from software specification

Abstract—Web services have become popular in the modern infrastructure of the World Wide Web. They aim to provide automatic discovery, selection, and invocation of required applications (services) across the internet. However, the quality assurance aspects of web services remain a challenge. Recently, the semantic web has been introduced as an emerging technology which emphasizes presenting the meaning of the web content to achieve a machine processable automation. In this paper, < Final Year Projects > we explore the synergy of applying specification based software testing techniques to semantic web services. Our approach investigates the possibility of deriving concrete test cases from the goal specification of a semantic web service in order to determine the correctness of a service implementation. Furthermore, we also propose coverage criteria to evaluate the generated test cases at both the goal and the service description levels. We demonstrate the generation and evaluation of the test cases from a goal specification with the help of a simplified discount example.