Using procedure reasoning system for knowledge discovery in data mining

Many techniques of knowledge discovery rely solely on attributes and their relationships that exist in databases. In contrast, the knowledge to find specific knowledge used in expert systems usually comes from human experience. To be beneficial for developing data mining technology, we introduce a procedure reasoning system that allows one to discover knowledge by combining human experience with data values and data patterns. In this paper, we emphasize system architecture of the procedure reasoning and decision-rules generation. The procedures of organizing and controlling the knowledge discovery process are discussed.