Research On Rough-Fuzzy Set Theory-Based Decision Support Approach Of Component Adaptation

In component-based software developing, it is difficult to directly reuse components. That system developers need to adapt components to meet the system requirements is called "component adaptation". In real system developing, system developers understand and adapt components subjectively. And adaptation schema selected by developers' knowledge, understanding ability and reuse experiences, without scientific and objective basis. Since the component adaptation schema decided by system developer has great influence on components' functions, behaviors and integration efficiency, it is the problem should be solved to enhance objectivity in adaptation decision and to increase the efficiency of component adaptation.Furthermore, although evaluation and feedback mechanisms used nowadays can summarize and compare the history reusing information of component, the results of analyze mainly focus on the quality and reusing history of the component, cannot evaluate the conformity degree of the component in current application, and cannot support the component adaptation decision well.According to problems mentioned above, in this paper we improved the traditional evaluation and feedback mechanisms. Firstly we improved on traditional component library architecture to propose a hierarchical component library architecture. And then, we proposed a rough-fuzzy set theory-based decision support approach of component adaptation. What's more, we gave the approach's algorithm. The approach can analyze and mine the feedback information more, thus well supports objective component adaptation decision and decreasing the cost to understand and adapt components.To evaluate the new approach objectively, we implemented the approach in our prototype component library for a case study. Consequently, we found that the approach can objectively figure out strategic scenario to improve the efficiency and accuracy of component reusing.