An efficient particle swarm intelligence based strategy to generate optimum test data in T-way testing

Limited resources and tight deadline factor inhibits exhaustive testing. Thus, generation of optimal test data in an acceptable number is very important to accelerate the overall software engineering process. Search based optimization technique has been used in software test data generation since 1992 with recently increasing interest and activity within this area. Brief literature shows that, a change to the parameter interaction (t-way interaction) can significantly reduce the number of test data. Based on this principle, many t-way test data generation strategies have been developed over the past decade. Recent finding state that, implementation of artificial intelligence based searching for test data generation can obtain near optimum solution. However, producing the optimum test data appear to be NP-hard problem (Non-deterministic polynomial). As such, it is almost impossible for a strategy to produce the optimal set of test data. With the analysis of recent studies of the valid different search based optimization approach, this paper represents a swarm intelligent based searching strategy to generate near optimum test data. The performances are analyzed and compared to other well-known strategies. Empirical result shows that the proposed strategy is highly acceptable in terms of the test data size.