International Journal for Development of Computer Science & Technology (IJDCST)
Poor quality data is a rising and expensive problem that affects many enterprises across all aspects of their business ranging from operational effectiveness to revenue protection. A novel type of semantic rules extended from traditional functional dependencies is proposed as Conditional Functional Dependencies (CFDs). In this paper, for detecting inconsistencies in data, the authors present an approach that efficiently and robustly discovers conditional functional dependencies and improves data quality. An expensive process that involves intensive manual effort is to find the quality of CFDs.