Functional Dependency (FD) is an important feature for referenc ing to the relationship between attributes and candidate keys in tuples. It also sh ows the relationship between entities in a data model (Calvanese et al. 2001). In research areas of data cleaning (Arenas et al. 1999; Bohannon et al. 2005), the FD is used for improving the data quality. In a data mining research, an FD discovery technique has been studied (Savnik and Flach 1993; Huhtala et al. 1999) . However, an FD discove ry could find too many FDs and, if use directly in a cl eaning process, could cause it to NP time (Bohannon et al. 2005). In this research, we have developed a cleaning engine by combining an FD discovery technique with data cleaning technique and use the feature in query optimization called “Selectivity Value” to decrease the number of discovered FDs. Testing results showed that this work can identi fy duplicates and anomalies with high recall and low false positive.