A Reduction Algorithm Meeting Users Requirements

Generally a database encompasses various kinds ofknowledge and is shared by many users. Different users may preferdifferent kinds of knowledge. So it is important for a data miningalgorithm to output specific knowledge according to users' currentrequirements (preference). We call this kind of data miningrequirement-oriented knowledge discovery (ROKD). When the rough set theoryis used in data mining, the ROKD problem is how tofind a reduct and corresponding rules interesting for the user. Sincereducts and rules are generated in the same way, this paper onlyconcerns with how to find a particular reduct.The user's requirement is described by an order of attributes, calledattribute order, which implies the importance of attributes for theuser. In the order, more important attributes are located before lessimportant ones. Then the problem becomes how to find a reduct includingthose attributes anterior in the attribute order. An approach todealing with such a problem is proposed. And its completeness forreduct is proved. After that, three kinds of attribute order aredeveloped to describe various user requirements.