Rearch of optimal reduct under preference

摘要

With more and more users sharing with the same data resource, different users have different preference and requirement to the same data information. Based on this, a new data mining technique emerges, called requirement oriented knowledge discovery. According to the problem of retrieving reducts in rough sets, this article presents a special preference relation - dictionary relation. The article describes the optimal reduct under such preference relation, constructs the data structure to solve the problem and the corresponding algorithm on the designed structure. At last it constructs the algorithm to retrieve the optimal reduct under the such preference relation, and proves the correctness of the algorithm theoretically.