Return to Article Details An Analysis on Attribute Selection and Token Formation used for Duplicate Record Detection