Dr Yashir,

Thanks for the extremely interesting talk yesterday. Your view on why Semantic Web is still struggling to be adopted was right on the mark, hopefully most of us in non-academic world will cherry pick some of the semantic concepts and adopt them in our apps.


Here are some links on Oracle 11g's Semantic technology Oracle Semantic Technologies Overview


rdf support <RDF support in Oracle 11g>

- Atul

I thought it was a very good talk yesterday, however I wonder why the meetup and talk were titled "Geospatial Semantics", yet that topic was not covered in even the smallest detail? Sorry, we did some geo data about a birthday cake :)

I was just hoping for a more on topic discussion as it is not always easy to get away for a few hours at night so I try to make those hours count.

Oops, I just changed the thread title to correct Dr. Bishr's name, but in doing so I inadvertently removed the Oracle link. Could you give the link again, Atul?

Thanks for contributing to the message board!

Yaser,

Which of the ontologies do you think is most mature, and will be the "VHS" rather than the Betamax of geospatial semantics?

You mentioned an OWL version of GML under development, can you give us a link and tell the group a little more about it? Is it worth learning more about?

Brian

I thought it was a very good talk yesterday, however I wonder why the meetup and talk were titled "Geospatial Semantics", yet that topic was not covered in even the smallest detail? Sorry, we did some geo data about a birthday cake :)

I was just hoping for a more on topic discussion as it is not always easy to get away for a few hours at night so I try to make those hours count.

Thanks for your feedback Frank. Like you, we are all taking away from our personal time to network and learn something new. That was my motivation to drive 130 miles last night. I will be glad to answer specific questions about geospatial semantics in this thread.

Yaser,

Which of the ontologies do you think is most mature, and will be the "VHS" rather than the Betamax of geospatial semantics?
You mentioned an OWL version of GML under development, can you give us a link and tell the group a little more about it? Is it worth learning more about?

SIOC, SKOS, RDFa, FOAF, SCOT, and the Dublin Core ontologies are the ones that have the chance of become widely adopted. They are all simple yet powerful and answer real need.
OWL GML is located (here http://efe.ege.edu.tr.... This is the only one that I'm aware of. This was developed a while ago. Is this ontology useful? It depends.

• It does not include the feature model, which is necessary.
  
• It does not include coordinate reference system
  
• If you are going to run this ontology through a reasoner, then it is useless. This is because it is OWL Full, which is undecidable by design. This means that there is never an answer to some queries, even if the answer really exists. To build real world applications, I wouldn’t recommend anything beyond OWL light at this point.
  
• If this ontology is just a model, like UML, then it is useful
  

Beyond an extremely narrow group within the hardcore GIS’ers, I have yet to see a real world application that would require this complexity. For the rest of us we need ontology that captures the spatial relations, and to a lesser extent geometries. Things like, within, intersects, overlaps. On top of this ontology we can build naive geography concepts like, far, close, next to, neighbor, etc.

A good practice when you build geospatial ontology, or any ontology, is to know what kind of inferences you will need, and accordingly you design your ontology in such a way that would let you infer the result you’d except to get from the inference engine accurately and FAST.

Best
Yaser

I will be glad to answer specific questions about geospatial semantics in this thread.

Again, thank you for taking the time to speak with us. I hope you did not take what I had said as a reflection of my opinions on your talk and efforts.

With respect to Geospatial Semantics, since you have done work in that arena, I was wondering what are some of the various ontologies for defining geo data, and if there are any that are considered to be "front runners" or more popular for providing this data as RDF. My frame of mind is in terms of usability and simplicity over absolute completeness. Like you had said, I want to implement this *now* and simplicy, in my mind, will make it more usable.

Thanks again,
Frank

With respect to Geospatial Semantics, since you have done work in that arena, I was wondering what are some of the various ontologies for defining geo data, and if there are any that are considered to be "front runners" or more popular for providing this data as RDF. My frame of mind is in terms of usability and simplicity over absolute completeness. Like you had said, I want to implement this *now* and simplicy, in my mind, will make it more usable.
Frank

Hi Frank,
I'm not aware of the existence of what i call naiive geography ontology. By that I mean spatial and temporal concepts and relations that people use in their everyday life. Something like nextTo, within, closeTo, farFrom, rightSide, leftSide, etc. Although these terms are used by almost every person, expressing them semantically is not easy. Mainly because spatial relations are always context dependant. For example, when I say, My house is on the rightSide. Is this egoCentric where I'm the reference or is it exoCentric where another object is the reference.

Actually your question led me to consider providing this ontology to the larger community. Give me sometime and I will get back to you on this one.

Best

Yaser