Key equivalence in heterogeneous databases

Abstract

Semantic heterogeneity in heterogeneous databases (HDBs) has been recognized as a difficult problem. One of the probable causes of the little progress made in this area is the lack of well-defined subproblems that individual researchers can attack and achieve results. The author formulates the key equivalence problem in HDBs. By its nature there is no single algorithm that can solve the general key equivalence problem, but the author proposes a framework (a meta-algorithm) to approach it. He applies the framework to a concrete instance of the problem, the name matching problem and describes his experience so far.<>