In this article we define a formalism to discover knowledge in the form of temporal rules, inferred from databases of events with a temporal dimension. The theoretical framework is based on first-order temporal logic and allows the definition of the main temporal data mining notions (event, temporal rule, constraint) in a formal way. The concepts of consistent linear time structure and general interpretation are funda-mentals in the design of algorithms for inferring higher order temporal rules, (called temporal meta-rules), from local sets of temporal rules.
展开▼