Artificial Bee Colony based Test Data Generation for Data-Flow Testing

Objectives: It is a challenging task to generate and identify an optimal test set that satisfies a robust adequacy criterion, like data flow testing. A number of heuristic and meta-heuristics algorithms like GA, PSO have been applied to optimize the Test Data Generation (TDG) problem. The aim of this research work is to handle the automatic Test Data Generation problem. Methods/Statistical Analysis: This research work focuses on the application of Artificial Bee Colony (ABC) algorithm guided by a novel Fitness Function (FF) for TDG problem. The construction of FF based on the concept of dominance relations, weighted branch distance for ABC to guide the search direction. Ten well known academic programs were taken for experimental analysis. The proposed algorithm is implemented in C environment. Findings: To examine the effectiveness of ABC algorithm in Test Data Generation, ten academic programs were taken experiment. The effectiveness of proposed algorithm is evaluated using average number of generations and coverage percentages achieve parameters. The experimental results show that proposed ABC algorithm requires less number of generations in comparison to other algorithms. It is also noted that the proposed algorithm coverage almost all def-path for all programs. Application/Improvements: The experimental results depict that the ABC algorithm performs far better than other existing algorithm for optimizing test data.