Abstract:

A conditional functional dependency (CFD) is an extension of the traditional functional dependency (FD). By introducing the conditional pattern, the CFD is more accurate and more expressive than FDs in semantics. However, it is time-consuming for computing CFDs. In order to improve the efficiency of CFDs, the incremental maintenance method for CFDs is studied. The changing rules on the conditions of three different situations (i.e., dataset insertion, deletion, update) are analyzed, and an incremental algorithm for calculating CFDs is proposed, so that we can efficiently and dynamically maintain CFDs while the database is changing. At the same time, the correctness of key steps of the algorithm is demonstrated, and the validness of the algorithm is verified through the experiments.