Test data for Record Linkage

Description

These tables contain artificial personal data for the evaluation of Record Linkage procedures. Some records have been duplicated with randomly generated errors. RLdata500 contains 50 duplicates, RLdata10000 1000 duplicates.

Usage

data(RLdata500)
data(RLdata10000)

Format

RLdata500 and RLdata10000 are character matrices with 500 and 10000 records. Each row represents one record, with the following columns:

fname_c1

First name, first component

fname_c2

First name, second component

lname_c1

Last name, first component

lname_c2

Last name, second component

by

Year of birth

bm

Month of birth

bd

Day of birth

identity.RLdata500 and identity.RLdata10000 are integer vectors representing the true record ids of the two data sets. Two records are duplicates, if and only if their corresponding values in the identity vector agree.

Author(s)

Andreas Borg

Source

Generated with the data generation component of Febrl (Freely Extensible Biomedical Record Linkage), version 0.3 (https://sourceforge.net/projects/febrl/). The following data sources were used (all relate to Germany):

http://blog.beliebte-vornamen.de/2009/02/prozentuale-anteile-2008/, a list of the frequencies of the 20 most popular female names in 2008.

http://www.beliebte-vornamen.de/760-alle_jahre.htm, a list of the 100 most popular first names since 1890. The frequencies found in the source above were extrapolated to fit this list.

http://www.ahnenforschung-in-stormarn.de/geneal/nachnamen_100.htm, a list of the 100 most frequent family names with frequencies.

Age distribution as of Dec 31st, 2008, statistics of Statistisches Bundesamt Deutschland, taken from the GENESIS database (https://www-genesis.destatis.de/genesis/online/logon).

Web links as of October 2009.