Optimal test sequence generation in state based testing using moth flame optimization algorithm

Abstract

Software testing contributes a strategic role in software development, as it underrates the cost of software development. Software testing can be categorized as: testing via code or white box testing, testing via specification or black box and testing via UML models. To minimize the issues associated with object-oriented software testing, testing via UML models is used. It is a procedure which derives test paths from a Unified Modelling Language (UML) model which describes the functional aspects of Software Under Test (SUT). Thus, test cases have been produced in the design phase itself, which then reduces the corresponding cost and effort of software development. This early discovery of faults makes the life of software developer much easier. Also, there is a strong need to optimize the generated test cases. The main goal of optimization is to spawn reduced and unique test cases. To accomplish the same, in this research, a nature-inspired meta-heuristic, Moth Flame Optimization Algorithm has been offered for model based testing of software based on object orientation. Also, the generated test cases have been compared with already explored meta-heuristics, namely, Firefly Algorithm and Ant Colony Optimization Algorithm. The outcomes infer that for large object-oriented software application, Moth Flame Optimization Algorithm creates optimized test cases as equated to other algorithms.