A knowledge-based approach for generating test scenarios for web applications

The popularity and complexity of current Web applications continue to increase, and thus, their quality and reliability are increasingly becoming more important. Software testing is an important part of validating the software, but it needs a large number of good quality test scenarios and test cases. This can be done manually, but it would be very time consuming. They can also be generated automatically, but previous work only considers artifacts of the software-under-test. Given that there are already a huge number of Web applications that have gone through rigorous testing, we believe that the information from these previous tests can be a good source of information for testing future Web applications. This paper proposes an approach to automatically produce test scenarios by leveraging on the knowledge gathered from existing Web applications that have already undergone software testing. Our approach is based on a database containing test scenarios that are collected from previous applications. When a tester wants to produce test scenarios for a new application, he/she first creates a "base" scenario of the application, and then searches the database for related scenarios. The retrieved scenarios are then used to test the applications. A case study showed that our approach can retrieve useful test scenarios that a professional tester was not able to build manually.