Useful functionalities for record linkage - ARCHIVED

Articles and reports: 11-522-X201300014270

Description:

There is a wide range of character-string comparators in the record-linkage field. Comparison problems arise when factors affect the composition of the strings (for example, the use of a nickname instead of a given name, and typographical errors). In these cases, more sophisticated comparators must be used. Such tools help to reduce the number of potentially missed links. Unfortunately, some of the gains may be false links. In order to improve the matches, three sophisticated string comparators were developed; they are described in this paper. They are the Lachance comparator and its derivatives, the multi-word comparator and the multi-type comparator. This set of tools is currently available in a deterministic record-linkage prototype known as MixMatch. This application can use prior knowledge to reduce the volume of false links generated during matching. This paper also proposes a link-strength indicator.

Issue Number: 2013000
Author(s): Lachance, Martin
FormatRelease dateMore information
PDFOctober 31, 2014

Related information

Subjects and keywords

Subjects

Date modified: