An Algorithm of Mining Classification Rules Based on Faster Constructing Formal Concepts

A novel algorithm called Fccrm is proposed,in which the dividing method is used to reduce the scale of formal context and the technology is adopted that gets all the lattice nodes by seeking all biggest concepts and their lower covers to avoid the complicated relations in concept lattice.The pre-pruning strategy is applied in the mining process to reduce the size of the formal concepts further,for getting the classification rules from the formal concepts.The experimental results show that the performance of this algorithm is obviously improved,and the classification rules set obtained from formal concepts is complete for the sample set.