Classification-tree restructuring methodologies: A new perspective

Abstract

The classification-tree method developed by Grochtmann et al. provided a useful approach for constructing test cases from functional specifications. It was automated by Chen and Poon through their tree construction methodology. In a follow-up study, Chen and Poon found that the effectiveness of constructing legitimate test cases could be improved under certain circumstances via a classification-tree restructuring algorithm. We develop another tree restructuring algorithm to cater for other situations not covered previously. The algorithms complement each other. We also compare the relative effectiveness of these algorithms and provide guidelines on applying them in practice.