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

Table 1 Some illustrative, fictitious, personally-identified records

Record Number	Name	Sex	Street address	Locality	Age in years
0	Gwen Palfree	F	Flat 17 23–25 Knitting Street	West Wishbone 2987 New South Wales	42
1	Angie Tantivitiyapitak	F	Wat Paknam Saint George Ave	Old Putney NSW 2345	27
2	Gwendolynne Palfrey	2	17/23 Knitting St	Wishbone West NSW 2987	42
3	Tontiveetiyapitak, Angela	Female	C/- Paknam Monastery, 245 St George St	Putney 2345	28
4	Palfrey, Lyn	1	Corner of Knitting and Cro	chet Streets, Wishbone New Sth Wales	43

Note: The "bleeding" of street address data into the locality column in record 4 is deliberate, and typical of real-life data captured by information systems with fixed-length data fields.

ISSN: 1472-6947