A formal framework for Data Mining process model
Ding Pan · 2009
Data mining is a dynamic research and development area that is reaching maturity, so it requires well-defined foundations, which are well understood throughout the community. The CRISP-DM process model seems to have become the dominant. A novel model for data mining is proposed in evolving environment, for continuous data mining. As the basis of the model, a formal framework for data mining and knowledge management is proposed to define main notions used in data mining in first-order linear temporal logic. It represents a rule in quasi-Horn clause, defines the measures of the first-order formula valuating on a linear state structure, and generates the estimator sequence of the measures based on a session model.