An Empirical Study of Regression Test Suite Reduction Using MHBG_TCS Tool

Software testing happens to be an inescapable, big-budget and protracted software engineering activity. It is well established that software testing cannot be accomplished in totality even for small input programs. The researchers have been trying to find alternatives like test case selection, minimization, and prioritization. Along with many other approaches, recent research has witnessed the emergence of many Search-Based Software Testing (SBST) techniques. One such technique for Regression Test Selection using evolved approach for hybrid use of Genetic Algorithms (GA) and Bee Colony Optimization (BCO) (MHBG_TCS) was proposed and empirically validated. This paper empirically attempts to find out the effect of varying Time Constraint (TC) values on the results obtained using this MHBG_TCS technique. 17 test programs were tested for 7 varying TC values. The results inculcate very high confidence in the optimality and efficiency of the technique for higher TC. The findings strengthen the scope of further research and use of MHBG_TCS technique in software testing.