![]() ![]() ![]() |
![]() |
|
|
![]() ![]() ![]() ![]() ![]() |
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
Return to Data Mining and Temporal Constraints We consider the problem of discovering temporal integrity constraints satisfied by a particular temporal database instance. We restrict our study to a particular class of temporal constraints called temporal algebraic dependencies which can be specified by simple temporal logic formulas and alternatively by pairs of algebra relational expressions or SQL-expressions. We argue that this class of temporal constraints capture the semantics of the most important temporal regularities currently found in real-world situations. We develop a technique of knowledge discovery in temporal databases which is based on a genetic programming method. Finally, we discuss implementation issues of the evolutionary algorithm Temp-Gen that has been developped for this problem and present some preliminary results. DiSC'01 Copyright ©2002 ACM Inc. |