AN IMPROVED APPROACH TO ADAPTIVE TESTING

Abstract

Adaptive testing is the counterpart of adaptive control in software testing. It means that software testing strategy should be adjusted on-line by using the testing data collected during software testing as our understanding of the software under test is improved. Previous studies on adaptive testing involved a simplified Controlled Markov Chain (CMC) model for software testing which employs several unrealistic assumptions. In this paper we propose a new adaptive software testing approach in the context of an improved and namely, general CMC model which aims to eliminate such threats to validity. A set of more realistic basic assumptions on the software testing process is proposed and several unrealistic assumptions are replaced by less unrealistic assumptions. A new adaptive testing strategy based on the general CMC is developed and implemented. Mathematical simulations and experiments on real life software are conducted to demonstrate the effectiveness of the new strategy.