- From: Gary Berg-Cross <gbergcross@gmail.com>
- Date: Sat, 22 Nov 2008 11:51:46 -0500
- To: paola.dimaio@gmail.com
- Cc: public-xg-eiif <public-xg-eiif@w3.org>
Paola et al, > 2. so, are you suggesting that everything that comes in our 'where' > partition, is modelled to correspond entities modelled by the standards that > we chose to adhere to, and follows the corresponding naming convention? It will take a bit of time for me to develop an adequate proposal to the group on this that we can work through, but here are some of my initial thoughts on this in relation to the question above. I think the approach I argue for is general and can be used for other parts of our framework, but for me it is easier to start there with the Where dimension of the model. We are developing a framework that includes Where concepts and we recognize the need to develop a common vocabulary with more explicit (semantic) definitions. I don't expect that everything in even a large set of geospatial standards will cover the Where concepts we may come up with. For one thing we may look at some of the information vocabulary "bottom up" from info needs and info in current systems and there is no guarantee that voc in existing covers these or covers these with the meaning that we want to use. So I see the voc in the standards as a starting point and foundational, but not a rigid foundation. But any deviation from them should provide more benefit that loss. So it is not > simply, keep our stuff as it is, and map it to the standard? In some cases a standard may provide a primitive that we can use in defining a composite idea that better serves our need and matches what we see in EM systems. For example we may have a standard definition of "area" but not its sub-type "disaster area". So if we thinking of using a scheam to transmit the information and there is a schema that has a section for "area" we might propose defining an extension with that standard schema to accomodate the data elements (say definred with a formal RDF) that we add to describe a "disaster area". > 1. are the standards that you mention and suggest that we reference the > only/best 'where' standards? I think Carl on this list is also working on > some, and it would be good if could make this choice with the widest > possible angle (aiming for reach) and do not compromise our future choices I agree that Carl could add some useful ideas to this. He wasn't on the call, but perhaps is catching up on the discussion. Regards, Gary Berg-Cross On Fri, Nov 21, 2008 at 10:19 PM, <paola.dimaio@gmail.com> wrote: > Gary, > thanks a lot > > let's start from the where as you say > > two thoughts come to mind > > 1. are the standards that you mention and suggest that we reference the > only/best 'where' standards? I think Carl on this list is also working on > some, and it would be good if could make this choice with the widest > possible angle (aiming for reach) and do not compromise our future choices > > 2. so, are you suggesting that everything that comes in our 'where' > partition, is modelled to correspond entities modelled by the standards that > we chose to adhere to, and follows the corresponding naming convention? or, > simply, keep our stuff as it is, and map it to the standard? Somehow I think > that the latter is preferable, that is, to have a schema that can easily be > mapped/matched to 'any' standard. In a way I am looking for internal > 'completeness' and 'consistency' of representation, leaving the external > mapping to be > done on an ad hoc basis (as required) > > I am interested in what people think is best > > > PDM > > > > On Sat, Nov 22, 2008 at 12:07 AM, Gary Berg-Cross <gbergcross@gmail.com> > wrote: >> >> Paola et al, >> >> My questions was about the role of this section "Towards Common >> Ontology" in the overall EII Framework. >> >> We say the role of the ontologies is for"collection of the semantics" >> and critique NIEM in not having an overall framework for the >> overlapping/ diverse semantics they include, but we provide no overall >> framework for semantics ourselves or point to one that we agree with. >> So my question was "what do we want to do in this section and how far >> would we try to go to frame a semantic approach here?" >> >> A common ontology by itself may not be a frame, but the way that >> ontologies are used to support our conceptual framework etc. might be >> the way this section contributes. >> >> It seemed in our discussion the group is starting to be more explicit >> and expressive about "emergency information" vocabulary - the who, >> what, where and when of emergencies. I think that this may be >> compatible with some of the other semantic interoperability thrusts >> and semantic annotation efforts. >> >> I offered to do a bit ofo review for the "where" dimension. Here's >> an example of what I was thinking of there. It's in 2 steps - >> leverage existing standards that have a start on a core vocabulary and >> then add some semantic rigor to this to better support >> interchangeability (integrate disparate information to present in a >> common form) and interoperability (use information at an application, >> pragmatic, and process control level). >> >> Here is an example of step one: >> >> Core vocabulary for Where (geospatial concepts) >> >> I would start with the Geography Markup Language (GML) - a standard >> whose model was originally based on the W3C Resource Description >> Framework (RDF). For exchange the OGC introduced XML schemas into >> GML's structure to help connect the various existing geographic >> databases, whose relational structure XML schemas more easily define. >> The resulting XML-schema-based GML retains many features of RDF, >> including the idea of child elements as properties of the parent >> object (RDFS) and the use of remote property references. >> >> For our where Q GML contains a rich set of primitives which are used >> to build application specific schemas or application languages. These >> primitives include: >> >> • Feature – a distinction from a geometry object. A feature is an >> object in our doaim that represents a physical entity, e.g. a >> building, a river, rescue area, or a person. We are primarily >> interested in these, but need the other concepts to locate them or >> describe them as in locating and describing a rescue area. >> >> • Geometry – things like Point. LineString or Polygon that may >> describe a Feature >> • Coordinate Reference System to provide coordinates of geometry >> objects (e.g. line coordinates) >> >> • Time (BTW for better semantics there is Jerry Hobb's OWL-Time >> http://www.w3.org/TR/owl-time/ >> • Coverage (including geographic images) >> • Unit of measure etc. >> >> This is a start we can leverage and they are also moving to enhance >> the semantics by "annotating" this vocabulary with better semantics. >> For example, there are many types of geospatial "part" relations that >> can be make more explicit. We should leverage this work and not try >> to re-invent it. >> >> Part of the work I am thinking of is out of the sensor standards work >> partly supported by NIST and by OGC. Amit Sheth has a demo on semantic >> annotations that enhance primarily syntactic XML-based descriptions >> with what OGC calls Sensor Web Enablement (SWE ) languages - >> microformats, and W3C's Semantic Web languages- RDF and OWL. The >> combination of semantic annotation and semantic web capabilities >> i(ontologies and rules) "supports interoperability, analysis and >> reasoning over heterogeneous multi-modal sensor data". >> >> http://knoesis.wright.edu/research/semsci/application_domain/sem_senso, >> July-August 2008, pp. 78-83. >> Amit Sheth and Matthew Perry, "Traveling the Semantic Web through >> Space, Time and Theme >> >> >> Gary Berg-Cross >> >> On Fri, Nov 21, 2008 at 6:59 AM, <paola.dimaio@gmail.com> wrote: >> > Gary, and all >> > >> > can you you please repeat the question that you had in commenting the >> > paragraph 'toward common ontology'? >> > >> > When I first joined this group, the word ontology itself was avoided, to >> > avoid entering in the realm of the ' too abstract and complicated' to be >> > useful. Obviously members of this incubator have become more comfortable >> > using the term since we started and, inevitably now we have to become >> > semantically more precise. >> > >> > So, if we have to talk about ontology in our draft framework document, >> > this >> > would be a good time to start >> > doing so properly. >> > >> > At the moment what we have is a schema, which in itself it would be >> > great >> > to have, because ultimately, in a functinal sense, that's what we need >> > to >> > make the information flow a bit more coherent , functional and >> > efficient. >> > >> > To make sure that our schema is compatible with the grand scheme of >> > things, >> > and universals and primitives, may require some additional refinement of >> > our >> > conceptual model. This will result in our schema to be more versatile >> > robust >> > and consistent and much more useful in time. >> > >> > Please share your thoughts with us, and let's ponder what choices we >> > have to >> > confront to move our work up to the metaphysical ladder (hehe, joking) >> > >> > Guido Vetere who has recently joined this group said that he is going to >> > send some considerations and suggestions on how to model our schema to >> > comply with DOLCE, >> > >> > http://www.loa-cnr.it/Papers/DOLCE2.1-FOL.pdf >> > >> > which I very much look forward to seeing his contribution >> > >> > >> > Gary, which foundational ontology have you worked with before? What >> > would be >> > your suggestions to >> > align our work with top level categories of sorts? >> > >> > I think it is a challenge for bottom up schemas (what we are doing now) >> > to >> > comply with foundational requirements, >> > as well as it is a chellenge for foundational ontologies to be >> > adopted/applied in bottom up schemas creation >> > >> > So starting thinking in terms of ontology proper is an interesting and >> > important exercise that we cannot longer postpone >> > and hopefully we'll learn what we need to learn along the way >> > >> > >> > pdm >> > >> > >> > >> > >> > >> > >> > >> > -- >> > Paola Di Maio >> > School of IT >> > MFU.ac.th >> > ********************************************* >> > >> > > > > > -- > Paola Di Maio > School of IT > MFU.ac.th > ********************************************* > >
Received on Saturday, 22 November 2008 16:52:23 UTC