[HOT] Matthew

john whelan jwhelan0112 at gmail.com
Wed Feb 22 15:30:16 UTC 2017

Data quality is always an issue and we don't have enough validators which
doesn't help.

We also have user expectations.  If you want to use OSM in a particular
area with low internet access perhaps you could arrange with someone to run
an eye over the area first?  With a large number of different mappers
mapping by hand I think you must expect some inconsistences. I've even seen
mappers with more than two years experience tagging in a way that didn't
follow local guidelines.

We know if we catch mappers making mistakes within the first 24 hours the
errors go down after that.  Crisis mapping is always going to be lots of
new people wanting to help.  In Nepal I think 70% of the mappers were first
time mappers and if you think data quality is bad in Haiti then Nepal was
off the scale.

Since the recent changes to the task manager, iD and the training guides I
have seen the number of errors decrease.

What I have noticed is some projects are better managed than others.  I
think a recent one I looked at the bottom 25% of the project had no mapable
imagery from Bing or other sources.  Other African projects have
instructions that do not follow the African highway wiki.  The problem here
is by not presenting a consistent professional image to the mappers it
gives the impression that we don't care, and that this stuff isn't very

Currently validation isn't anyone's responsibility.  Perhaps as part of the
project plan check list the project managers should be asked what plans do
they have to ensure that 50% of new mappers work will be validated within
48 hours or whatever the magic window Martin's work shows and that does not
include asking me to validate yet another project. I have my own
idiosyncrasies about which I validate on and it does take a fair bit of
time and effort to do it right.

I sympathise with the idea of commenting "Just clean up this mess!" but
recognise Blake's comment of pat them on the head and make them feel a
little happier is valid as well.  For the most part we are talking
volunteers so persuasion works best, having said that something very
interesting is taking place in Lusaka.  I'm fairly certain from the mapping
being done that although the mappers are new to OSM they do know GIS and
the quality of mapping is very high.  Project 2543 is an example.  There
has also been some work counting buildings etc using R and the local level
of government seem to be involved.

Cheerio John

On 22 February 2017 at 09:37, Fred Moine <frmoine at gmail.com> wrote:

> Sorry for that I was in mountain area with the student of UNOGA with
> limited acces to internet and time.
> If you do the a query over Grand Anse area http://osmose.openstreetmap.fr
> /fr/map/
> You will discover that is not possible to use OSM like that due to a
> number of duplicate building.
> So please include a strong validation or cleaning process after each
> activation.
> As is not acceptable to work like that and let all this mistake....
> All the best FredM
> 2017-02-22 15:08 GMT+01:00 Blake Girardot HOT/OSM <
> blake.girardot at hotosm.org>:
>> Dear Fred,
>> Thank you for the feedback.
>> It would help if you provided an actual area and not just an generic
>> link to osmose.
>> I did review all of NW Haiti for duplicate buildings which is what
>> seems to be the biggest issue and out of 68,000 buildings over
>> approximately 1500 sq/km found 470 duplicate buildings. ( graphic
>> https://screenpresso.com/=2eYrd )
>> Which is pretty good in my view, less than 0.7% error rate. Still it
>> would be good to fix them, especially since it is pretty easy to fix.
>> For an experienced mapper, it should take about 1 hour to fix them all.
>> I fixed 100 of them in about 15 mins the slow way :)
>> I would be happy to share my work flow for fixing them if that would
>> help you at all.
>> My only other comment is that if you are really interested in getting
>> the errors corrected, being a bit more understanding and respectful of
>> the volunteer mapper's efforts might motive people to dive in and
>> help. Phrases like "clean up your mess" is not a good approach in my
>> experience. "Hey, I found a lot of duplicate buildings, could we find
>> a way to get them corrected?" would probably have generated a lot more
>> volunteers to jump in and help.
>> Cheers,
>> Blake
>> On Wed, Feb 22, 2017 at 6:39 AM, FredM <frmoine at gmail.com> wrote:
>> > We are in training with student from Universite Nouvelle Grande Anse
>> > http://universitynouvellegrandanse.org/ and we are using drone et OSM.
>> >
>> > However, it is a lot of mistake done after the cyclone
>> >
>> > Have a look on osmose and check all  to see all duplicate building and
>> other
>> >
>> > http://osmose.openstreetmap.fr/fr/map/
>> >
>> > Not so funny to start cleaning an OSM database when you start the
>> mapping.
>> >
>> > All the best FredM
>> >
>> > ---
>> > L'absence de virus dans ce courrier électronique a été vérifiée par le
>> > logiciel antivirus Avast.
>> > https://www.avast.com/antivirus
>> >
>> >
>> > _______________________________________________
>> > HOT mailing list
>> > HOT at openstreetmap.org
>> > https://lists.openstreetmap.org/listinfo/hot
>> --
>> ----------------------------------------------------
>> Blake Girardot
>> Humanitarian OpenStreetMap Team, TM3 Project Manager
>> skype: jblakegirardot
>> HOT Core Team Contact: info at hotosm.org
>> Live OSM Mapper-Support channel - https://hotosm-slack.herokuapp.com/
>> BE A PART OF HOT'S MICRO GRANTS: https://donate.hotosm.org/
> --
> Moine Frédéric
> Potentiel3.0 / OSM member
> Haiti : 36 19 45 44
> mail: frmoine at gmail.com
> Skype: fmoine74
> http://www.potentiel3-0.org/index.php/en/
> _______________________________________________
> HOT mailing list
> HOT at openstreetmap.org
> https://lists.openstreetmap.org/listinfo/hot
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstreetmap.org/pipermail/hot/attachments/20170222/1d2c855d/attachment.html>

More information about the HOT mailing list