Fuzzy C Means (FCM) Clustering Based Hybrid Swarm Intelligence Algorithm for Test Case Optimization

The main objective of an operative testing strategy is the delivery of a reliable and quality oriented software product to the end user. Testing an application entirely from end to end is a time consuming and laborious process. Exhaustive testing utilizes a good chunk of the resources in a project for meticulous scrutiny to identify even a minor bug. A need to optimize the existing suite is highly recommended, with minimum resources and a shorter time span. To achieve this optimization in testing, a technique based on combining Artificial Bee Colony algorithm (ABC) integrated with Fuzzy C-Means (FCM) and Particle Swarm Optimization (PSO) is described here. The initiation is done with the ABC algorithm that consists of three phases-the employed bee, the onlooker bee and the scout bee phase. The artificial bees that are initialized in the ABC algorithm identify the nodes with the highest coverage. This results in the ABC algorithm generating an optimal number of test-cases, which are sufficient to cover the entire paths within the application. The node with the highest usage by a given test case is determined by the PSO algorithm. Based on the above ‘hybrid’ optimization approach of ABC and PSO algorithms, a set of test cases that are optimal are obtained by repeated pruning of the original set of test cases. The performance of the proposed method is evaluated and is compared with other optimization techniques to emphasize the fact of improved quality and reduced complexity.