On 6/16/05, Drew Taylor <[EMAIL PROTECTED]> wrote: > On Jun 16, 2005, at 7:51 PM, Joel Gwynn wrote: > > > Of course, until Google Maps is officially released, and until the get > > more specific about their TOS, this is all just having fun. > > Not to nitpick, but the Maps TOS basically says, "See our primary TOS" > - which explicitly forbids commercial usage.
*sigh* ... picky, picky :) > > It's funny you mentioned this, because at $WORK I'm faced with the > problem of geocoding a whole bunch of addresses for our database. And > since most searches on the database are radius searches, lat/long is > rather important. > > In our case, a lot of the addresses already have lat/long. But many of > them are incorrect, which is worse than missing! So I have about 75k > addresses that need to be redone. The few commercial services I've > looked at would be quite expensive. > > One option we've used in the past is just doing zipcode centroid > matching. You can get this information for ~$100. But obviously this is > less accurate. In my case, I'm not sure if the hit in accuracy is too > much. I need to do more checking. > > Has anyone used commercial services and been happy with the > price/results? In the case of using the TIGER/Line dataset, how > accurate is it? > > Drew I'm not working on commision for Eagle, or anything, but their prices seem pretty reasonable: 100k addresses for $1535.00 https://www.etakcentral.com/EZ_Locate/ez_subscribe.html Of course, in the handful of test cases I used, Geo::Coder::US was just as good, for free. In a couple of days you could set up your own geocoder and be all set. Note, Geo::Coder::US requires perl 5.8.x, which I did NOT have when I first installed it. The key is to realize that all data is bad, it's just a question of how much time/money you're willing to spend to make sure it's good. _______________________________________________ Boston-pm mailing list [email protected] http://mail.pm.org/mailman/listinfo/boston-pm
