[SensorML] sml:classifer: want multiple values--string, URN, URL

Luis Bermudez bermudez at sura.org
Tue Apr 21 09:19:58 EDT 2009


The code space should be then call context because it may seem
confusing. But yes it is important to have a context, which in the
semantic web world are commonly represented as "named graphs".

The point is that our semantic framework should be able to support the
following:
Lets say we have a concept C. A community A could say everything about
a concept C in their context. A community B could say everything they
want about concept C in their context. And we should be able to query
both concepts and get different information about C, even if they
contradict each other.

And in some cases and for practical purposes:  code space = context=
URI of ontology where the concept was define


-luis

___
Luis Bermudez Ph.D.
Coastal Research Technical Manager
Southeastern Universities Research Association (SURA)
bermudez at sura.org - Office: (202) 408-8211
1201 New York Ave. NW Suite 430, Washington DC 20005



On Tue, Apr 21, 2009 at 6:08 AM,  <Simon.Cox at csiro.au> wrote:
> I think there is still a role for the code space - it provides the context
> within which the concept is framed.
> AFAICT the same concept can be a member of more than one ontology.
> The context from which it is being used does supply more information,  in
> particular about what it is _not_.
>
> Simon
>
> ______
> Simon.Cox at csiro.au  CSIRO Exploration & Mining
> 26 Dick Perry Avenue, Kensington WA 6151
> PO Box 1130, Bentley WA 6102  AUSTRALIA
> T: +61 (0)8 6436 8639  Cell: +61 (0) 403 302 672
> Polycom PVX: 130.116.146.28
> <http://www.csiro.au>
>
> ABN: 41 687 119 230
>
>
>
> ________________________________
> From: John Graybeal [mailto:graybeal at mbari.org]
> Sent: Tuesday, 21 April 2009 12:24 PM
> To: Cox, Simon (E&M, Kensington)
> Cc: bermudez at sura.org; Jeff.deLaBeaujardiere at noaa.gov;
> sensorml at lists.opengeospatial.org; mike.botts at nsstc.uah.edu;
> Alexandre.Robin at spotimage.fr
> Subject: Re: [SensorML] sml:classifer: want multiple values--string, URN,
> URL
>
> By way of agreement with Simon... this convention (of code space being the
> base URI) also assumes the separator between the base and the term is known
> -- in the semantic community, it could be either / or #, if the separator is
> not included explicitly  in the code space.
> I recommend the full URI is the value, as Luis suggests.  But then what is
> the meaning or usefulness of the code space?  If it has not been defined
> already, perhaps Oceans IE can propose a best practice?
> john
>
> On Apr 20, 2009, at 9:07 PM, <Simon.Cox at csiro.au> wrote:
>
>> the code space is the ontology where this concept was define, which
>> usually is the base URI
>
> code-space = base URI is a common pattern in the Semantic Web community - so
> common that it is generally assumed.
> Its convenient for URLs, but may not match governance.
> In particular it makes it tricky for an item to be a member of more than one
> set.
> IMHO it is better to separate the concept of set and resource more clearly.
>
> Simon
>
> ______
> Simon.Cox at csiro.au  CSIRO Exploration & Mining
> 26 Dick Perry Avenue, Kensington WA 6151
> PO Box 1130, Bentley WA 6102  AUSTRALIA
> T: +61 (0)8 6436 8639  Cell: +61 (0) 403 302 672
> Polycom PVX: 130.116.146.28
> <http://www.csiro.au>
>
> ABN: 41 687 119 230
>
>
>
> ________________________________
> From: sensorml-bounces+simon.cox=csiro.au at lists.opengeospatial.org
> [mailto:sensorml-bounces+simon.cox=csiro.au at lists.opengeospatial.org] On
> Behalf Of Luis Bermudez
> Sent: Wednesday, 15 April 2009 4:41 AM
> To: Jeff deLaBeaujardiere
> Cc: sensorml at lists.opengeospatial.org; Mike Botts; Robin, Alexandre
> Subject: Re: [SensorML] sml:classifer: want multiple values--string, URN,
> URL
>
> Hi Jeff, et al
>
> Note, that what we want to know ( have identified in URI) is the type of
> system ( or process), which in this case is a mooredBuoy or a 6-meter NOMAD
> buoy
> so the sml:value should be a URI
>
>  e.g. http://mmisw.org/ont/mmi/20081116T072152/platform/MooredBuoy
>
> the code space is the ontology where this concept was define, which usually
> is the base URI (http://mmisw.org/ont/mmi/20081116T072152/platform
>
>
>
> -luis
>
> ___
> Luis Bermudez Ph.D.
> Coastal Research Technical Manager
> Southeastern Universities Research Association (SURA)
> bermudez at sura.org - Office: (202) 408-8211
> 1201 New York Ave. NW Suite 430, Washington DC 20005
>
>
> On Mon, Apr 13, 2009 at 5:39 PM, Jeff deLaBeaujardiere
> <Jeff.deLaBeaujardiere at noaa.gov> wrote:
>>
>> Mike, thank you for your answers to my earlier questions.
>> I am following through on your suggestion to use the SensorML
>> mailing list for additional questions like the one below.
>>
>> My present need is for metadata about a station (e.g., a buoy) that
>> contains several sensors.
>> Suppose I have "station type" as a classifier:
>>
>>  <sml:classifier name="Station Type">
>>   <sml:Term definition="urn:x-noaa:def:classifier:NOAA::stationType">
>>     <sml:codeSpace
>> xlink:href="http://www.ndbc.noaa.gov/mooredbuoy.shtml"/>
>>     <sml:value>6-meter NOMAD buoy</sml:value>
>>   </sml:Term>
>>  </sml:classifier>
>>
>> I would like to have multiple values:
>>
>> * The human-readable type such as '6-meter NOMAD buoy'
>> * An identifier like urn:x-noaa:def:stationType:NOAA::6N
>> * An explicit URL for metadata about this type:
>>
>>  http://.../sos?requst=DescribeSensor&procedure=urn:x-noaa:def:stationType:NOAA::6N
>>  (this URL is so I can factor out common stuff into a separate response,
>> and allow
>>  that response to be harvested by web crawlers that do not know how to
>> construct
>>  SOS DescribeSensor requests).
>>
>> How can I put multiple values in? The schema for sml:classifier and
>> children is too limiting (I think).
>> Has anybody else encountered this use case?
>>
>> Thanks
>>
>
> _______________________________________________
> SensorML mailing list
> SensorML at lists.opengeospatial.org
> https://lists.opengeospatial.org/mailman/listinfo/sensorml
>
>
> John
> --------------
> John Graybeal   <mailto:graybeal at mbari.org>  -- 831-775-1956
> Monterey Bay Aquarium Research Institute
> Marine Metadata Interoperability Project: http://marinemetadata.org
>
>



More information about the SensorML mailing list