Research of Electronic Patient Record Mining Based on Rough Concept Lattice

According to the features of the medical data in the electronic patient record database, the rough concept lattice mining algorithm (RCLM) is put forward. The condition entropy algorithm is adopted to reduce the attributions ,then concept lattice and variable accuracy rough set are related, thereby, a concept decision rule lattice is put forward. Which based on the prune policy in constructing algorithm, the lower approximation decision regular lattice is gained and the corresponding consistent decision rule is extracted from this decision regular lattice. Finally the mining experiment of the electronic patient record is designed. The results show that RCLM is better on both sides of the running speed and the mining capability.