Based on Attribute Order for Dynamic Attribute Reduction in the Incomplete Information System

As the real-world information system is constantly updated, knowledge acquisition based on the need of users has become an important issue in current data mining. Attribute order can reflect the user's needs and interests, which reflects the importance of attributes to different users. In this paper, we first make use of the extended rough sets model with limited tolerance relationships and provide a new information entropy function. Then we compute an attribute reduction in a dynamic incomplete decision system on the basic of the attribute order. Moreover, when we consider the case of adding or deleting an object by the incomplete information system, an incremental reduction algorithm and a reduced reduction algorithm are given. Experiments show that the feasibility and effectiveness of the proposed algorithm.