A Constructing Algorithm of Concept Lattice with Attribute Generalization Based on Cloud Models

Data mining is to extract the implicit, potential useful information from data. For containing the trivial detail of original concept hierarchies, a large number of rules are directly mined from relational databases, and most of such rules are unnecessary. However, mining information in high hierarchies will generate some rules that are interesting and useful. In this paper, cloud models is adopted to control the generalization of a set of qualitative attributes, and a new constructing algorithm of concept lattice based on cloud models is presented for mining association rules in large databases, which is integrating attribute-oriented generalization and concept lattice. Finally, a specified experiment is conducted to illustrate the superiority of rule knowledge discovery using such lattice with attribute generalization