Comparison of metahuristic test generation strategies based on interaction elements coverage criterion

Abstract

Interaction testing represents an important technique, among those broader testing techniques involved in the software test data generation process. Within this technique, test cases are selected using a combination of test input parameters. We normally want that all combinations of parameters' values (called interaction elements) occur in the test suite at least once. Metaheuristics search algorithms have been used for constructing an interaction test suite by constructing test cases that can cover all the interaction elements. This paper introduces the interaction elements coverage as criterion to compare different metaheuristic interaction test suites generation strategies. In doing so, this paper gives an extensive review for different metaheuristic test generation strategies. The comparison results shows that by using the particle swarm optimization more interaction elements can be covered with fewer test cases and iterations.