Date Added: Aug 2012
Conditional Functional Dependencies (CFDs) are the extension of Functional Dependencies (FDs) by supporting patterns of semantically related constants. These CFDs are very useful to frame the data cleaning rules in relational databases. The CFDs have been proven more effective than FDs in detecting and repairing dirtiness of data. However, finding the CFDs is a difficult task and it also involves rigorous manual effort. This paper proposes a fast and efficient method called Best CFD to find the CFDs from relations. This algorithm works effectively find the CFDs compared to the other algorithms like CTANE. The proposed method is implemented in Java with SQL database connectivity and tested for different datasets. The results validate the effectiveness of this method for finding CFDs in databases.