[ML4OSM] Measure machine learning assisted mapping

[Bo Percival]

Some great looking work there Felix, it's going to be great to see this
empirical approach going forward. Follow the evidence wherever it leads!

@Flo, you make a strong point on the community establishment. I have a
couple of follow up questions for you, for the sake of a conversation by
which we can facilitate a lot of learning for others.
1. I really liked the examples you use here of communities that may have
been negatively impacted by imports. Noting the Tiger data event, do you
think the community in the US was stronger or weaker after this happened?
And how/why? Also, can you just confirm for me, was the Tiger data AI
predicted?
2. You cite some very strong examples, however they are only really
representative of well developed countries, can you think of any countries
that are less developed that may have positively impacted from this kind of
activity in their country (excluding Thailand, as I wouldn't really
consider it 'less developed' at the time of the import)
3. I'd be really interested in also hearing your opinions on how you think
it could work. Do you think it would be possible to build communities with
advanced data techniques? Or do you think it is infinitly impossible? What
suggestions would you have to make it possible?

Thanks for taking the time to respond, this is such a great discussion
topic.

Le mar. 2 juin 2020 à 11:15, Florian Lohoff <f at zz.de> a écrit :

> On Tue, Jun 02, 2020 at 07:24:54AM +0000, Felix Delattre via
> machine-learning wrote:
> > Dear all,
> >
> > As you probably know, the conversation about machine learning techniques
> > and their use for OpenStreetMap has been very emotional in our
> > community. Opinions range from the potential negative impacts this could
> > have, to the hope that it would significantly improve the quality and
> > also the speed of OSM mapping, because it allows people to focus on what
> > they do best.
> >
> > We want to take an evidence-based look at the effects of machine
> > learning mapping on OpenStreetMap. To do this we are working together
> > with several organizations (German Geoscience Research Center,
> > University of Heidelberg and the OpenStreetMap humanitarian team) to
> > conduct research that will quantify the measurable impact of the
> > currently proposed mapping workflow. We believe that a reproducible and
> > transparent study will give us a clue.
> >
> > We are planning to do an experiment comparing four different datasets
> > from the same area:
>
> I put another one in the Basket.
>
>         "Long term community establishment."
>
> If you'd follow the last decade of OSM you might have noticed
> that we had countries with huge dataset imports, and we had countries
> with nothing, everything hand-mapped.
>
> If you look 10 years later the ones with huge imports struggle to
> establish a sustainable community to fix the imports,
> and maintain the data. (As a hint - Tiger Data US import, and AND import
> of the Netherlands)
>
> And from an long-term OSM point imports have been largely seen
> as a problematic issue for long-term community maintenance of data.
>
> Flo
> --
> Florian Lohoff                                                 f at zz.de
>         UTF-8 Test: The 🐈 ran after a 🐁, but the 🐁 ran away
> --
> machine-learning mailing list
> machine-learning at openstreetmap.org
> https://lists.openstreetmap.org/listinfo/machine-learning
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstreetmap.org/pipermail/machine-learning/attachments/20200602/545ed244/attachment.htm>