Algorithm of Hierarchical Reduction Based on Rough Entropy

We propose a scheme based on rough entropy to cluster the concept hierarchy for resolving the problem of discovering knowledge from large databases. The paper defines rough entropy and establishes the relation between the information system knowledge and rough entropy. It offers a new algorithm based on rough entropy and attribute significance to construct information. The attributes are classified to different parts allocated at several layers, so the knowledge in the information system can be presented hierarchically with multiple granularities at multiple layers. The time complexity of the algorithm is analyzed and the algorithm is used to a control decision, which verifies the effectiveness