Orientation-based Ant colony algorithm for synthesizing the test scenarios in UML activity diagram

Abstract

Abstract Context The model-based analysis is preferred over the code-based analysis as it speeds up the development process and directs the guiding effort. In the software industry, the Unified Modeling Language (UML) is a set standard followed by the developers as well as system analysts to extract all attainable paths of controls, usually known as scenarios under an activity diagram. Objective In this manuscript, a bio-inspired methodology has been applied on concurrent sub-part of a UML activity diagram to fetch various feasible test scenarios. Method The food search pattern of an ant has been taken as a base heuristic. An orientation factor has been introduced in the existing ant colony optimization algorithm. Experiments have been performed using three student projects, five synthetic models and an openly available model repository named LINDHOLMEN data-set at Github. Results The statistical analysis has validated the results obtained through various existing approaches and the proposed approach. Experimentation shows that the orientation-based ant colony algorithm has produced better results as compared to the existing Genetic Algorithm (GA) and Ant Colony Optimization (ACO) on the basis of feasible test scenarios generated.