Token-based method of blocking records for large data warehouse

Record linkage is a critical problem in duplicate data elimination. It is used to detect and eliminateduplicate data. The elimination of duplicate data will increase the quality of data. Record Linkage problem willtake high computational cost because of the large number of record comparisons. The comparison of records isinefficient in large data warehouses. Blocking methods are used to group the records to minimize the number ofrecord comparisons. This paper explains the existing blocking methods and its comparison and discusses theselection of token-based blocking key for record comparisons.