Hello Everybody,

Is there a way to Geocode European addresses (unparsed data) using Mappoint?

Sample Data format:

SCHYTTTESVEJ 7 8700 HORSENS, Denmark
GRENVÄGEN 10 14171 HUDDINGE, Sweden

From simple Google search 8700/14171 doesn't seem like zipcodes.

Thanks!
Uday

Hello,

Our MapPoint-Tool IC-AddressValidation will validate and geocode addresses. Addresses of (most) European-countries and addresses of USA and Canada.
Look at the manual(WEBHELP) : IC-AddressValidation

We geocoded your 2 addresses in Sveden and Denmark.
Just have a look to the screenshots.

If you, or somebody else, is interested to get a trial version, please send us an email.

mail@profit100.de

Kind reagards
Manfred


PROFIT1oo™ Consulting - Microsoft® MapPoint ISV (Independent Software Vendor)
Internet: http://www.mappoint2010.info , MS MapPoint - Geoanalyse - Vertriebsgebiete - Routenplanung - MapPoint Training - MapPoint Schulung

In this case it's easy to separate the address, postcode, city, country just looking at the spaces, you could use Excel String functions, but in cases where the address has an additional space or the city has two words, this could get complicated.

SCHYTTTESVEJ 7 8700 HORSENS, Denmark
GRENVÄGEN 10 14171 HUDDINGE, Sweden

Can you post more of your list?

Eric

Thanks Manfred and Eric for your suggestions. Sorry for the delay in replying to the thread.

Eric - As you suggested I have been trying to parse the data using Excel and map it using mappoint, but as you guessed due to the nature of the data there was lot of massaging involved which is so time consuming. Thats why I wanted to see if there is a way to map unparsed data.

At the moment I have decided to take the excel route and map the data, but one of the challenges I came across is Mappoint's ability to cluster cities to Counties/Provinces. For instance if my input has $100 revenue for all the cities mapped in Sweden, when I cluster the data by Provinces mappoint shows only $80 revenue.

Did anyone come across this issue?


Uday

If working through the API, then a useful trick is to try various variations until you get a 'hit'.

I had this problem with a load of Scottish data when I first started coding MapPoint years ago. UK addresses are generally not as structured as US ones (which all the examples use!), so there can be legitimate variations in the city/other city/region fields. also the data could be badly formatted (we were using local gov housing association data - ie. the land owners, but they were still inconsistent in their database formatting!).

So for example if you have 5 fields. The first is probably street. But is the second one "City", "Other City" or "Region"? Try all variations with the most likely first.

That sounds like a good idea. I will try that see how the match rate looks.


Uday