Abstract: Rough set theory is a new mathematical tool to deal with imprecise, incomplete and inconsistent data. The traditional rough set theory can only deal with the discrete attributes in database. However, most real-life databases consist of not only discrete attributes but also continuous attributes. In order to overcome the limitation of the traditional rough sets, using feedback information from decision table consistency we propose a new method of discretization of continuous attributes based on dynamic layer cluster. A unified framework of the rough set theory to deal with discrete and continuous attributes is suggested, which extends the scope of application of rough sets. The results of comparison between this method and some existing algorithms of discretization are encouraging.