Balanced Iterative Reduction and Clustering Using Hierarchies (BIRCH) methods to Prioritize and reduce Test Suites

Abstract

Throughout testing, there is an abundance of duplicate test cases to ensure that the new code will not impact the program that has previously been tested with every possible combination. Essential as it may be, software testing may be expensive, especially when test cases are not based on real-world events. By combining state-of-the-art Balanced Iterative Reducing and Clustering Using Hierarchies (BIRCH) techniques with the Fruit Fly optimization (FFO) methodology, we can use fewer test cases to reduce challenging time while still providing accurate results and making effective use of resources. BIRCH does away with the original data points and instead creates clustered summaries from them. The BIRCH-FFO clustering method produced a full test suite with full path coverage while significantly reducing the number of experimental cases.