Adaptive random testing by static partitioning

Abstract

Despite the importance of the random testing approach, random testing is not used in isolation, but plays a core role in many testing methods. On the basis that evenly distributed test cases are more likely to reveal non-point pattern failure regions, various Adaptive Random Testing (ART) methods have been proposed. Many of these methods use a variety of distance calculations, with corresponding computational overhead, newly proposed methods like ART by bisection, random partitioning or dynamic iterative partitioning try to decrease computational overhead while maintaining the performance. In this article we have proposed a new ART method that has similar performance to existing ART methods while having less computational overhead.