Object-Oriented Software Testing and Maintenance Methodology

Funded by:

The object-oriented paradigm (OO) is widely adopted by numerous Texas firms. However, this new paradigm also creates understanding, complex dependencies, and tool support problems in the OO software testing and maintenance phases. In the last three years, UTA has identified, along with researchers elsewhere, these problems and developed an OO test model and algorithms to solve some of these problems.

The project will 1) extend the test model to support polymorphism and dynamic binding, two important OO features that are challenging to testers and maintainers; 2) develop innovative methods and efficient algorithms for black-box testing and template class testing; 3) enhance the test model generation algorithm to recover complex, interative, and recursive design patterns; 4) refine and validate the test criteria and metrics by applying them to large OO programs; and 5) conduct controlled experiments to assess the usefulness and effectiveness of the results.

The project will advance the OO technology by offering effective and efficient solutions to the most challenging OO testing and maintenance problems. It is expected that the research will drastically reduce OO testing and maintenance costs and significantly improve OO software quality.

UTA's research on OO testing and maintenance is internationally recognized through more than 50 publications in ACM/IEEE and other international journals and conferences. To date, over $2,000,000 research grant has been received for OO testing. Supporting this project will significantly enhance the technology base and the infrastructure of science and engineering.

The OOTWorks tool set is now being evaluated at IBM, Raytheon, Ericsson and Fujitsu for commercial release. Response from industrial collaborators has been very positive. Please contact

Dr. David Kung
(817) 272-3627

for further information.