The calculation of Variable precision explicit region is an improved algorithm for constructing decision tree on the use of variable precision rough set model. For defects in the process of calculating the explicit region—in the process of calculating the explicit region, the more the number of attributes is, the greater the value of specific areas is, it puts forward the calculation algorithm that the number of attributes limits specific area. This algorithm enhances the accuracy of the calculation process. It can effectively reduce the trend that the more the classification of attributes is, the greater the greater the value of specific areas is. In the meanwhile, it also effectively improves the accuracy of the algorithm. By introducing the support and confidence, it simplifies the resulted tree, and improves the generalization ability of the tree. Finally, the validity of the method is verified through experimental analysis.