[OSM-talk] Adding wikidata tags to the remaining objects with only wikipedia tag

Stefan Keller sfkeller at gmail.com
Mon Oct 2 10:45:55 UTC 2017


Hi Yuri and all

Just my 2 cents:

I think the reasoning of Lester makes sense: be careful about
semi-automatic adding wikidata tags to any OSM object with a Wikipedia
tag.

Applied to a subset of OSM this already took place and now there are
some objects remaining, which will need to be curated by hand.
Specifically we're speaking about 750 remaining OSM objects with wrong
Wikidata tags (disambig. instead of "real" resources).
And that's a feasible number to be done by hand, isn't it?

Just for the ground truth following input:
* Wikipedia tags are used in several services, apps and tools, like
Nominatim(!), OSMNames.org, historic.places, etc..
* Wikidata tags are used by Mapbox, OpenMapTiles, Mapzen(?), etc.

So both tags have already their place.

In this thread it's Yuri just asked this initially:
> I would like to auto-add all the corresponding wikidata based on wikipedia,
> for all remaining objects, using  JOSM's "Fetch Wikidata IDs".

Pls. correct me if I'm missing something here. Though "auto-add" is
perhaps not the best notion, it's still a human in control of JOSM.
To me, OSM really should remain a welcoming, inclusive do-ocracy.
So, let's look forward.

:Stefan

P.S. To anybody eager to debate, IMHO the following OSM Wiki pages
need some care, and this is something really useful to discuss and
update:
* https://wiki.openstreetmap.org/wiki/Key:wikipedia
* https://wiki.openstreetmap.org/wiki/Key:wikidata


2017-10-02 10:22 GMT+02:00 Lester Caine <lester at lsces.co.uk>:
> On 19/09/17 21:03, Yuri Astrakhan wrote:
>> There is now a relatively small number of OSM nodes and relations
>> remaining, that have wikipedia, but do not have wikidata tags. iD editor
>> already automatically adds wikidata to all new edits, so finishing up
>> the rest automatically seems like a good thing to do, as that will allow
>> many new quality control queries. I would like to auto-add all the
>> corresponding wikidata based on wikipedia, for all remaining objects,
>> using  JOSM's "Fetch Wikidata IDs".
>
> Yuri
> I'm going to wind things back in a bit here as the discussion seems to
> be widely disjointed.
>
> I think there needs to be a formal discussion as to just what secondary
> data sources are preferred when adding additonal data to OSM objects.
> Wikipedia is a useful secondary source for a vast range of material, but
> some objects in OSM will not be 'notable' enough for wikipedia to allow
> an article to exist, so an alternative mechanism is needed for those
> objects. wikidata may well be a suitable alternative, but the simple
> fact that a more detailed article for a wikidata object by not be
> accepted makes wikidata something of a problem as well. While
> wikipedia/wikidata provide a sort of standard framework for additional
> data, the primary link from an OSM object should perhaps be to websites
> specific to the object rather than the filtered wikipedia view of the
> world?
>
> Some of the tangential debate has been re ADDING links to OSM object
> that have not yet been tagged with something suitable, and this is where
> cross matching these items depends on the information already available
> on in the OSM tags. I have still to be convinced that wikidata is
> 'independent' enough to be a reliable source of cross-reference links,
> especially where other reference tags are already used. UK Schools we
> have added the reference from the schools database. I've not looked to
> see if a wikidata 'view' of that data is available, but I would not look
> to add wikidata id's in addition to the school id's - BUT the wikipedia
> reference has been added where the schools are notable enough to warrant
> an article. I WOULD only look to tidying these objects to ADD the
> wikidata id if one was also checking that all three elements are
> correct, rather than simply automatically adding them. Whilst I was
> processing that data during the UK group push on it there were often a
> lot of corrections made to get the right 'set' of data on a school
> object. While only a small number of objects were actually wrong, that
> is enough to justify needing a manual cross check of some sort.
>
> --
> Lester Caine - G8HFL
> -----------------------------
> Contact - http://lsces.co.uk/wiki/?page=contact
> L.S.Caine Electronic Services - http://lsces.co.uk
> EnquirySolve - http://enquirysolve.com/
> Model Engineers Digital Workshop - http://medw.co.uk
> Rainbow Digital Media - http://rainbowdigitalmedia.co.uk
>
> _______________________________________________
> talk mailing list
> talk at openstreetmap.org
> https://lists.openstreetmap.org/listinfo/talk



More information about the talk mailing list