Automated Test Execution Effort Estimation Based on Functional Test Specifications

A usual activity performed to ensure quality is software testing. Although important, testing is an expensive activity. For that reason, test teams should be able to estimate the required effort to perform their test activities on the schedule and to request more resources or negotiate deadlines when necessary. This is even more important when regarding regression testing, where the execution of hundred of tests may be required to verify if a software application still working after some change in the code. In order to estimate the effort to execute the tests of a given test suite, we proposed an automated effort estimation model that regards test size and execution complexity measured from functional test specifications written in natural language. This paper describes a tool that supports the measurement of test size and execution complexity, as well as the estimation of test execution effort. The tool was designed to be extended according to the customers needs, such as different estimation models, file formats of test specifications, parsers APIs, etc.