Rough Set reduction is a typical iterative process; however, the user cannot give extra controls and preferences during the consecutively iterative process. In this paper, we propose a novel approach to Rough Set reduction by using control science viewpoint. In this model, information system is regarded as controlled plant, user’s preference about attributes is regarded as control objective, and the iterative algorithm designing process is regarded as control law designing. We have investigated the properties of Rough Set reduction based on control approach, and have designed the control system based on the properties, where single attribute set and user specified attributes are all used as core attributes to control a pruning process, and other core attributes worked out by previous steps are also used, iteratively. Such that it forms a dynamic closed-loop control by which the user can give much more interactivities. We have also implemented the experimental platform, and the experimental results show that the proposed approach is efficient and effective.
