Tim Churches wrote:
ash wrote:
Horst Herb wrote:
On Tuesday 24 October 2006 20:13, Tim Churches wrote:
500,000 records and have gone to 1 million), we now use the US Bureau of
the Census files at http://www.census.gov/genealogy/www/freqnames.html -
these have univariate frequencies associated with each name so that
probability-based selection can be used (albeit probabilities and names
wonderful!
And I have the Aussie postcode database from
http://www1.auspost.com.au/postcodes/index.asp?sub=2
I guess I can generate "random" birthdates programmatically and even
probabilistically weigh their frequency according to Aussie demographics
What's lacking now is a street database ...
Any ideas where to get such data from?
There are about 600 real street names in the generate.py utility I
mentioned, I think.
can you harvest them from white pages ?
Those street names might have well been laboriously harvested from the
White Pages. In the past one could buy an electronic version of the
White Pages for just a few hundred dollars, before such things were
taken off the market due to changes in the telecommunications
regulations and copyright challenges by Telstra (and a good thing too,
as they were the stuff of telemarketers' dreams)
so an old one would be fine for horst's purposes - anyone kept one ?
ash
_______________________________________________
Gpcg_talk mailing list
[email protected]
http://ozdocit.org/cgi-bin/mailman/listinfo/gpcg_talk
