Research on Formalism of Temporal Data Mining

A fundamental structure for temporal data mining is proposed, based on first-order linear temporal logic, to define the main notion used in temporal knowledge discovery in a formal way. The formalism does not impose what search and optimization method to use to discover temporal rules. The relevant properties of the notion are proved. The concept of linear time structure allows associating each time moment with an assignation of all symbols of a restricted first-order language L, and the measures about the extent of truth of a formula. According to the notions of the mining slice, a process model of continuous data mining is developed