[Imports] [Talk-us] Boston, MA, USA addr:housenumber Import

Roman Yepishev rye at keypressure.com
Sun Mar 20 18:52:11 UTC 2016


Hi, all! Going back to the point...

I stopped linking to the .osn files in order to minimize the confusion.

Instead, every neighborhood gets their .gpx file with the details of
the issues encountered at that particular address:
 - Building is missing.
 - Building has more than one address.
 - Building has a street name that is not known to OSM or has not been
added to the mapping.

This will allow locating the issues quicker while surveying.

I have also generated the .pbf file containing the last export of US,
MA from Geofabrik + unique house numbers. Since I am using OsmAnd, I
have also generated the .obf map. Basically, that's how OSM would look
like if the current list of changes is applied.

https://wiki.openstreetmap.org/wiki/Import/Boston_Street_Address_Manage
ment_%28SAM%29_Import#OSM_Data_Files

There was a question of the data accuracy earlier - I've spent the last
3 days spot-checking the building numbers and I have to say they are
pretty accurate. Some buildings, however, have picked one number from
the range assigned, and this will have to be rectified manually as
well. The data from Boston Tax Parcel may help with that, but it's
license is not completely clear.

I found a couple of issues while generating the datasets/surveying the
locations:
 * Some of the addresses from MassGIS point to parks, monuments, or
buildings are located a few feet away. While it may possible to assign
the marker to the building nearby, I'd prefer not to do the guess work
and leave it for refining in the future.
 * Original OSM street names must have been imported from MassGIS
roads, and a few of these don't correspond to the signs (something as
minor as Conry Crescent Street vs Conry Crescent or slightly more
interesting Glenvale Terrace vs Chestnut Terrace). So there are
clusters of buildings in .gpx file pointing to possible street naming
issues, which will have to be manually correlated with the signs on the
streets.
 * Multipolygon buildings are not supported by my script, so there will
be up to 48 'missing building' false positives for the whole Boston.

Overall, I feel pretty good about the quality of the data and would
like to hear more about issues that I may have missed.

-- 
Sincerely,
Roman Yepishev
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 801 bytes
Desc: This is a digitally signed message part
URL: <http://lists.openstreetmap.org/pipermail/imports/attachments/20160320/9184d495/attachment.sig>


More information about the Imports mailing list