Re: Ontology for geometric data?

Hi all,


Yes, geometry is out of scope for a Building community group. Yet, we do have a lot of complex 3D geometries in our areas that we would want to describe (often going way beyond what is in the geospatial domain). I can imagine that there is a lot of requirement for complex 3D geometry also in other areas, such as product development areas (cars, planes, chairs, ...). Many of these more complex 3D geometries rely on STEP, I think.


I think it would be of value to conceive a group that works on diverse geometric data models over the web (plenty of kinds), outside of the existing groups.


The GEOM ontology that was mentioned in the TPAC is a derivative from the STEP geometry model within the Industry Foundation Classes (a file-based standard in buildingSMART). It is just one of the possible examples. The presentation also refers to the use of Well-Known Text and .obj-based strings.


kind regards,

Pieter


--
prof. dr. ir.-arch. Pieter Pauwels
Department of Architecture and Urban Planning, Ghent University, Belgium
________________________________
Van: Frans Knibbe <frans.knibbe@geodan.nl>
Verzonden: vrijdag 26 oktober 2018 13:11
Aan: maxime.lefrancois@emse.fr
CC: public-lbd@w3.org; public-sdwig@w3.org
Onderwerp: Re: Ontology for geometric data?

Hi Maxime,

Threads are getting entangled now, I see. Earlier I posted this message<https://lists.w3.org/Archives/Public/public-sdwig/2018Oct/0090.html> on the SDWIG public list, trying to argue that a general ontology for spatial data should be developed.

I too believe that a geometry ontology goes beyond the scope of the LBDCG. But it does seem to be in scope for the SDWIG. That said, it was also in scope for its predecessor, the SDWWG<https://www.w3.org/2015/spatial/wiki/Main_Page>, but in the end no work on the subject was done. That probably was due to a lack of available time for the group members, and because of the heavy focus on geography within the group. There was the idea to work on an improved version of GeoSPARQL, but to my knowledge that has not started yet.

Perhaps it will helpful to know that the LBDCG is interested in a general ontology for geometry. I think other interest groups in the W3C sphere should also be interested in the topic, for instance those concerned with web graphics (2D or 3D) and the Web of Things.

An aside: On slide 46 ("External ontologies for 2D & 3D geometry") of the TPAC presentation an ontology with the name "Geom" is mentioned. Would that be http://data.ign.fr/def/geometrie?

Regards,
Frans




Op vr 26 okt. 2018 om 11:37 schreef Maxime Lefrançois <maxime.lefrancois@emse.fr<mailto:maxime.lefrancois@emse.fr>>:
Dear Frans,

Your question to the Linked Building Data Community Group mailing list is also highly relevant to the Spatial Data on the Web Interest Group, so I forward your mail to their public list to get some feedback as well.

What is currently defined in the Building Topology Ontology document (one of the reports of the LBD-CG) is here: https://w3c-lbd-cg.github.io/bot/#3DModel
As for working on a 3D geometry ontology, it was decided that this goes beyond the scope of the Linked Building Data group and shall be left out for some more domain-independent group to work on.
See also our presentation of LBD at TPAC 2018 this monday https://w3c-lbd-cg.github.io/lbd/presentations/out/TPAC2018.pptx slides 44-46.

The SDW IG recently published (while it was a W3C Working Group) some recommendations and notes including:
- Spatial Data on the Web best practices https://www.w3.org/TR/sdw-bp/
- Semantic Sensor Network Ontology https://www.w3.org/TR/vocab-ssn/
- OWL-Time Ontology https://www.w3.org/TR/owl-time/
You can get a quick overview of what they are currently investigating on their GitHub projects webpage: https://github.com/w3c/sdw/projects

Best,
Maxime

Le ven. 26 oct. 2018 à 11:19, Frans Knibbe <frans.knibbe@geodan.nl<mailto:frans.knibbe@geodan.nl>> a écrit :
Hello,

I have recently finished a small research project in which BIM data from an IFC file were to be combined with geographical (GIS) data. A Linked Data based approach was also investigated, So it was an example of something that should be made possible by Linked Data: combining data from different domains. A known but difficult problem arises when this is tried: the information models and data formats for geometry in BIM and GIS are fundamentally different. Even when GIS and BIM data are made available as Linked Data, their geometric data are not interoperable.

So my question is: Is anyone aware of an attempt to have more general (domain independent) specifications of spatial data on the semantic web? Or is this perhaps something that this group has already set out to do?

Regards,
Frans