Using OpenStreetMap Data to verify extractions

Hello,

I am working on commercial project where we extract data out of scanned documents.

To verify the result of the senders street name I would like to check the extraction result with a database of street names.
So when I extract “Sonnenallee” I want to check if there is a street name in the database with that name. If so, I am more confident that the extraction result is correct.
When I extract “Sonnena//ee” because of an OCR error I would like to check the database if there is a street name that is similar to my extraction result and suggest to use the entry in the database instead of the extracted result.

Is this possible with openstreetmap and allowed to use the data in such a way?

Thanks for your help :slight_smile:

Best,
Georg

Hi Georg,

it depends for which area you want to extract the names, for a specific city, for whole Germany or for the whole world. It may also depend on your level of skill and the operating system you use.

A solution which works under Linux is described in this posting. The thread is rather old but the solution should still work as described, maybe with some slight adaptations. Of course it also works with any other extract from Geofabrik.

Rainer

Hi Rainer,

Thanks for you response!

It looks very promising. Do you also know if it is allowed to use the data from Geofabrik for such a commercial project?

Best,
Georg

If you are creating a cleaned up database, it will be a derivative of OSM, and therefore, if you publish it at all, it will need to be under an OSM compatible licence.

I think it is generally accepted, in the OSM community that you can use other maps to find errors, but you must correct them from free sources, so using OSM to flag an error for manual intervention would be OK.

Internal use of corrected data is also generally acceptable, but you mustn’t, for example, then sell a resulting mailing list.

OSM encourages commercial use, but does not encourage the production of proprietary databases.