From: Peter Christen <peter.christen@an...> - 2002-12-23 00:58:06
The Febrl team would like to inform you that two new publications
on name and address cleaning and standardisation as implemented
in Febrl are available online.
* Preparation of name and address data for record linkage using
hidden Markov models.
Published online at BioMed Central Medical Informatics and
* Probabilistic Name and Address Cleaning and Standardisation
Australasian Data Mining Workshop, Canberra, December 2002.
(go to Publications section)
Since the initial release of Febrl 0.1 in September, the Febrl team
has been working on an object-oriented re-design of the Febrl system,
making it easier to configure and modify.
Routines for traditional Fellegi and Sunter linkage and deduplication
(including blocking) are now included in Febrl.
The new version of Febrl (0.2) is currently being tested at the NSW
Department of Health and a release can be expected sometime in February 2003.
We wish you all a Merry Christmas and a Happy New Year.
Peter Christen and Tim Churches
Principal Developers of Febrl