A Software Decoupling Partition Method Based on Interactive Genetic Algorithm

AbstractWith the development of software component technology, extraction of components from legacy systems has become a reasearch focus of both software reuse and program comprehension for its lower cost and high efficiency. In this paper, we analyze the independence metrics of component for object-oriented software system from the perspective of coupling, cohesion and scale. Then, we propose a software partition decoupling method based on interactive genetic algorithm. By the user adjusting individuals fitness value, reduce the crossover and mutation probability, our method in obtaining reasonable component partition result while avoiding algorithm premature convergence