[OHM] Linked Data

Sat Apr 11 14:04:08 UTC 2015

Example RDF output(schema links are dummy links), lots of work has to be
done to make it play nice with RDF namespaces... This is not written
against Overpass, that would require more work.

https://gist.github.com/Abbe98/6f485ee8a0b421401dc6

2015-04-11 12:36 GMT+02:00 Albin Larsson <albin.post at gmail.com>:

> Tod,
>
> Let me change my statement,  "exposing raw RDF is not end-user friendly".
>
> The things with triple tags is that they are supported by the OSM-platform
> out of the box.
>
> When querying a OHM element you has to parse tags, using a triple tag
> system for relations allows you to do the same for relations. RDF as a
> value or tag forces you to implement both tag parsing and RDF/RDF data
> models(Such as OWL and EDM).
>
> //
> Albin
> On Apr 11, 2015 12:44 AM, "todd.d.robbins at gmail.com" <
> todd.d.robbins at gmail.com> wrote:
>
>> I wouldn't go so far as to say that "RDF is not end-user friendly". I
>> think that's on the UI/UX failures up to this point. The subject +
>> predicate + object model can be displayed and UI elements designed in a way
>> that makes the representation of relationships easy to document. Autofills
>> with short notations, for instance, solve a lot of the ambiguities of
>> deciding "hmm same_as vs. is_instance_of". I guess what I'm trying to say
>> is that the unfriendliness of RDF to an end-user isn't necessarily or
>> primarily because of the model for assertions but the UIs that have been
>> attempted thus far by non-designers. Also, I think Albin is on the right
>> track by investigating the triple tag/machine tag format (
>> https://en.wikipedia.org/wiki/Tag_(metadata)#Triple_tags) for expressing
>> complex relationships.
>>
>> Cheers!
>>
>> –Tod
>>
>>
>>
>> On Fri, Apr 10, 2015 at 3:50 PM, Albin Larsson <albin.post at gmail.com>
>> wrote:
>>
>>> I would like to go with the ohm:uri:same_as and a
>>> ohm:uri:is_instance_of, the idea it self is based on the idea that a mapper
>>> or end user will never see a format such as RDF or JSONLD.
>>>
>>> Your first example would be a is_instance_of.
>>>
>>> As I wrote earlier we should support RDF/other formats, but not by force
>>> the mapper to use them, ohm:uri:same_as would be equal to owl:sameAs(I
>>> think(but has to look into it)) so developers would be translating the
>>> tagging to RDF based on a schema we should provide. Then by creating some
>>> middleman software and a basic API we could provide RDF/JSONLD/... output
>>> and give developers a easier life.
>>>
>>> Would take maybe a week of work to create such a API.
>>>
>>> So short story, owl:sameAs has a equal relation tag, that tag just has
>>> to be translated. This is done because RDF is not end-user friendly. Try
>>> finding a place to enter RDF at Wikidata...
>>>
>>> //
>>> Albin
>>>
>>> 2015-04-10 22:30 GMT+02:00 Rob H Warren <warren at muninn-project.org>:
>>>
>>>> Albin,
>>>>
>>>> owl:sameAs would allow us to link the object in OHM space to other
>>>> databases, such as DB/Wikipedia/WikiData:
>>>>
>>>> Linking the OHM version of say
>>>> http://en.wikipedia.org/wiki/Roman_aqueduct
>>>>
>>>> Or linking the ww1 trenches within OHM to their Muninn equivalent.
>>>>
>>>> Or linking greek structures with their  pelagios equivalent
>>>> http://pelagios-project.blogspot.ca/
>>>>
>>>> Or linking modern administrative locations with their geonames,org
>>>> location.
>>>>
>>>> Since a LOD version of OHM will be positioned to be the equivalent to
>>>> dbpedia in historical GIS terms, the use of owl:sameAs would enable people
>>>> to discover non-OHM resources since it is the most obvious LOD data set to
>>>> link to.
>>>>
>>>> -rhw
>>>>
>>>>
>>>> > On Apr 6, 2015, at 10:58 AM, Albin Larsson <albin.post at gmail.com>
>>>> wrote:
>>>> >
>>>> > Sorry for the delay answering, I have been busy with other stuff...
>>>> >
>>>> > About the OGC idea I can't say more then that it would be devastating
>>>> to break the existing tools, the existing OHM instances(the rails-fork) is
>>>> hard enough to maintain.
>>>> >
>>>> > Rob could you explain future why owl:sameAs is needed and provide a
>>>> use case? I'm not getting the idea...
>>>> >
>>>> > //
>>>> > Albin
>>>> >
>>>> >
>>>> > 2015-04-02 15:55 GMT+02:00 Rob H Warren <warren at muninn-project.org>:
>>>> > Albin,
>>>> >
>>>> > I'd add owl:sameAs integration to the list of tags so that we can use
>>>> OHM as a resource discovery mechanism. -rhw
>>>> >
>>>> >
>>>> > > On Mar 27, 2015, at 4:12 PM, historic-request at openstreetmap.org
>>>> wrote:
>>>> > >
>>>> > > Date: Fri, 27 Mar 2015 13:57:36 +0100
>>>> > > From: Albin Larsson <albin.post at gmail.com>
>>>> > > To: "Historic at openstreetmap.org" <historic at openstreetmap.org>
>>>> > > Subject: [OHM] Linked Data
>>>> > > Message-ID:
>>>> > >       <CAM-QGEmn+WwHCK4eee24Nn=+rPvxjFdSLqJ5=fqS33m=
>>>> Dw1osQ at mail.gmail.com>
>>>> > > Content-Type: text/plain; charset="utf-8"
>>>> > >
>>>> > > My thoughts on linked data in OpenHistoricalMap and how I do it:
>>>> > >
>>>> > >
>>>> http://abbe98.github.io/blog/2015/03/26/mapping-the-past-with-linked-data-in-openhistoricalmap/
>>>> > >
>>>> > > Feedback, ideas, thoughts?
>>>> > >
>>>> > > //
>>>> > > Albin
>>>> >
>>>> >
>>>>
>>>>
>>>
>>> _______________________________________________
>>> Historic mailing list
>>> Historic at openstreetmap.org
>>> https://lists.openstreetmap.org/listinfo/historic
>>>
>>>
>>
>>
>> --
>> Tod Robbins
>> Digital Asset Manager, MLIS
>> todrobbins.com | @todrobbins <http://www.twitter.com/#!/todrobbins>
>>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstreetmap.org/pipermail/historic/attachments/20150411/7ab0fb57/attachment.html>