An Algorithm for Computing the Finite Closure of Attributes for Temporal Functional Dependencies with Multiple Time Granularities

In order to design temporal databases effectively, temporal functional dependencies (TFDs) supporting for multiple granularities of time are used to normalize temporal schemes. A crucial problem that needs to be solved for normalization of temporal databases is to compute the finite closure of attributes for TFDs. Usages of multiple granularities of time make it very complex to compute the finite closure of attributes. In fact, TFDs are proper extensions of traditional functional dependencies (FDs). Thus, there exist close relationships between TFDs and traditional FDs. By analyzing the relationships and properties of close sets of temporal types, an algorithm for computing effectively the finite closure of attributes is proposed by using the corresponding algorithm for traditional FDs. According to the analysis and the comparison with related algorithms, the algorithm proposed is more effective than the other ones.