Advanced Review Record linkage

This article describes methods for matching duplicates within or across files using non-unique identifiers such as first name, last name, date of birth, address, and other characteristics. 2010 John Wiley & Sons, Inc.WIREs Comp Stat Record linkage, in the present context, is simplythe bringing together of information from two records that are believed to relate to the same entity—for example, the same individual, the same family, or the same business. This might involve the linking of records within a single database to identify duplicate case records. Alternatively, record linkage might involve the linking of records across two or more databases. Such work might be undertaken to merge these databases into a single database with improved coverage or scope. The record linkage work is easiest when unique identification numbers (such as Social Security Numbers) are readily available. The work is more challenging when only quasi-identifiers such as given name, surname, date of birth, and address are available. In combination, quasi-identifiers may uniquely identify an individual. APPLICATIONS OF RECORD LINKAGE Record linkage is widely used by both businesses and government agencies. Businesses might use record linkage techniques to remove duplicate entries from mailing lists, thereby reducing both printing and mailing costs, and otherwise operating more efficiently. Businesses might also use record linkage to improve the functionality of their databases. For example, in the corporate combination between Bank of America and Countrywide Mortgage, it was widely conjectured that a major factor driving the deal was Bank of America’s desire to obtain e-mail This report is released to inform interested parties of research and to encourage discussion. The views expressed are those of the author