A hybrid weight-based and string distances using particle swarm optimization for prioritizing test cases

Regression testing is concerned with testing the modified version of software. However, to re-test entire test cases require significant cost and time. To reduce the cost and time, higher average percentage fault detection (APFD) rate and faster execution to kill fault mutant are required. Therefore, to achieve these two requirements, an improvement to existing Test Case Prioritization (TCP) technique for a more effective regression testing is offered. A weight-hybrid string distance technique and prioritization using particle swarm optimization (PSO) is proposed. Distance between test cases and weight for each test case, and hybridization of both values for weight-hybrid string distance are calculated. This experiment was evaluated using Siemens dataset. Result obtained from this experiment shows that weight-hybrid string distance is capable of improving APFD values whereby APFD value for hybrid TFIDF-JC is equal to 97.37%, which shows the highest improvement by 4.74% as compared to non-hybrid JC. Meanwhile, for percentage of test cases needed to kill 100% fault mutants, hybrid TFIDF-M yields the lowest value, 22.88%, which shows a 76% improvement as compared to its non-hybrid string distance