Abstract: Formal concepts analysison and rough sets theory are two different fast developing tools for data mining. The advantages of both the concept lattice in formal concepts representation and rough sets in knowledge reduction were cnough taken, and the algorithm of formal concepts rough mining(FCRM) under the incomplete electronic patient record knowledge system was put forward. The algorithm can carry on incomplete knowledge formal concept representation and rough positive approximate reduction with the decision rule lattice, and the corresponding consistent decision rule was extracted. Finally the expert system of the traditional Chinese medicine(TCM) patient record was designed. The experimental results show that algorithm of FCRM is better on the knowledge reduction and mining capability.

Key words: Incomplete knowledg, Formal concepts analysison, Rough positive approximate reduction, Electronic patient record mining, Decision rule lattice