For a while, I was geo-coding every address on CraigsList housing ads for a map mash-up.
I tested all the geocoders included in Geo.py: http://mail.python.org/pipermail/python-announce-list/2006-October/005297.html. This includes geocoder.us. I found that the only one that did the job even remotely reliably was Google, followed by Yahoo. Geocoder.us didn't seem to code very many addresses at all. The addresses people list on CraigsList on often very vague, so maybe this isn't a good test of the geo-coders, but I found Google was the only really good one. I guess there's some chance I wasn't using the others properly as well. Andrew Co-founder, TrailBehind.com On Thu, Jul 16, 2009 at 1:06 PM, Andrew Turner <[email protected]>wrote: > Christopher Schmidt wrote: > > On Thu, Jul 16, 2009 at 02:36:47PM -0400, Sean Gorman wrote: > > > Hi Fernando - > > The geocoder we just open sourced has address parsing in it that could do the > job: > http://github.com/geocommons/geocoder/tree/master > > Which may well be similar to: > > http://search.cpan.org/~sderle/Geo-StreetAddress-US-0.99/US.pm > <http://search.cpan.org/%7Esderle/Geo-StreetAddress-US-0.99/US.pm> > > (Which is the code from geocoder.us.) > > > > bigger, better, faster, and now with more peanut butter (as the glue). > > The goal is to extend parsers, and data importers for various countries' > addressing schemes and data sources (esp. OpenStreetMap). > > Andrew > > > > _______________________________________________ > Geowanking mailing list > [email protected] > http://geowanking.org/mailman/listinfo/geowanking_geowanking.org > >
_______________________________________________ Geowanking mailing list [email protected] http://geowanking.org/mailman/listinfo/geowanking_geowanking.org
