Preparation of name and address data for record linkage using hidden Markov models

Table 2 Segmented and transformed versions of the records from Table 1

Data element	Record 0	Record 1	Record 2	Record 3	Record 4
Given names	gwen	angie	gwendolynne	angela	lyn
Surnames	palfree	tantivitiyapitak	palfrey	tontiveetiyapitak	palfrey
Sex	female	female	female	female	male
Institution names		paknam		paknam
Institution types		monastery		monastery
Unit types	flat
Unit identifiers	17		17
Wayfare numbers	23,25		23	245
Wayfare names	knitting	saint george	knitting	saint george	knitting, crochet
Wayfare types	street	avenue	street	street	street
Wayfare qualifier					corner
Locality name	wishbone	putney	wishbone	putney	wishbone
Locality qualifiers	west	old	west
Territories	nsw	nsw	nsw		nsw
Postcodes	2987	2345	2987	2345

ISSN: 1472-6947