abstract
- In recent years, the increased complexity of Web applications gives researchers an enormous challenge in ensuring their reliability. A statistical testing approach has been proved to be appropriate to test Web applications and to estimate their reliability. Based on a Markov usage model built from user sessions, testers can generate abstract test cases randomly. In order to execute the test cases, researchers proposed an approach to building a data model from user sessions to provide test data for abstract test cases. However, just building a data model from all the user sessions may cause unauthorized access to restricted pages by testers. In this paper, we propose a concept analysis based approach to statistical Web testing and present a tool to support our approach. We have designed three algorithms to implement our concept analysis based approach that can provide the concrete test data for abstract test cases. In addition, a case study is introduced to demonstrate the effectiveness of our approach to statistical Web testing.