Research On Several Dynamic Renewal Methods Based On Granular Reduct

With the advent of the information age,information update is frequent and huge.If the information needed is re-mined through the updated formal context,the time cost will be greatly increased.Therefore,it is important to use the original knowledge to acquire the knowledge in new concept lattice.Therefore,in this paper,we discuss the updating rule of granular reduct in dynamic environment.Generally,there are three ways of data updating,including object updating,attribute updating and attribute value updating.The updating of attribute values can be regarded as deleting some objects(attributes)and then adding some objects(attributes).Therefore,it is only necessary to deal with the updating method of granular reduct when objects and attributes are gradually deleted or increased.Based on the above discussion,this paper mainly studies the updating rule of granular reduct in a dynamic formal context via the ideas of granular formation and granular transfer.The difference the method in this paper and the traditional method is that the traditional method directly calculate the granular reduct in updated formal lattice.This paper uses the idea of dynamic updating to calculate the granular reduction in updated formal lattice after analyzing the dynamic updating law of the granular reduction,which effectively avoid the repeated calculation.The main achievements and innovations are as follows.(1)The definition and properties of granular coordination set are given.(2)The corresponding updating method is proposed for dynamic updating of granular coordination set when an object or attribute is deleted.At the same time,the related algorithms are given,and some numerical experiments are conducted to show the effectiveness of the proposed algorithms.(3)The corresponding updating method is proposed for dynamic updating of granular coordination set when an object or attribute is added.At the same time,the related algorithms are given,and some numerical experiments are conducted to show the effectiveness of the proposed algorithms.