simon,
i've been chewing on this for a while. i'm not sure whether the DC community
(which is not all semantics community...., ok most of it) has more than
approximate solutions as to how we draw values from authoritative
dictionaries or sources. maybe stu weibel will comment on this. what we
should do however, is to compare notes on the ways it is done in our
respective communities.
if you could give me an example from your world, we could circle it through
the DC-General list and see what people 'll say about it.
best regards
thom
> -----Ursprüngliche Nachricht-----
> Von: Simon Cox [mailto:[log in to unmask]]
> Gesendet: Dienstag, 10. September 2002 03:08
> An: [log in to unmask]
> Betreff: Re: dc-environment application profile
>
>
> Thom - In the OGC Observations and Measurements model (which
> is based on VIM
> and NRC1995 and Fowler's analysis pattern for Observations)
> we "type" the
> observation in two ways:
>
> 1. Semantically - using a parameter "observable" or "measurand"
> 2. Structurally - in terms of the datatype: quantity, category, etc.
>
> The "observable" will have a value like "Temperature" or
> "Water Quality"
> which is drawn from an authoritative dictionary or source, in
> which it is
> defined using prose, or better still a language that supports
> Knowledge
> Representation or at least ontology. We only have
> approximate solutions to
> how this would be done, and would **welcome engagement from
> the semantics
> community** in resolving this.
>
> Note that in many cases the result of an observation is
> actually a composite
> value, so the description of both the datatype and the
> observable may need
> to be able to represent the structure used to serialise the composite.
>
> Also, some semantic observables need to be refined with parameters -
> OrganismCount will usually be refined in terms of its
> "species"; Radiance
> will usually be refined in terms of the frequency band.
> These parameters
> are actually just new typed values - all very recursive.
>
> _____
> [This mail represents part of a discussion of work in progress
> and should not be used for any purpose without my permission.]
> _____
> [log in to unmask] CSIRO Exploration & Mining
> 26 Dick Perry Avenue, Kensington WA 6151
> PO Box 1130, Bentley WA 6102 AUSTRALIA
> T: +61 (8) 6436 8639 F: +61 (8) 6436 8555 C: +61 (4) 0330 2672
> http://www.csiro.au/page.asp?type=resume&id=CoxSimon
>
> > -----Original Message-----
> > From: Automatic digest processor [mailto:[log in to unmask]]
> > Sent: Tuesday, 10 September 2002 7:20 AM
> > To: Recipients of DC-ENVIRONMENT digests
> > Subject: DC-ENVIRONMENT Digest - 3 Sep 2002 to 9 Sep 2002 (#2002-20)
> >
> >
> > There is one message totalling 296 lines in this issue.
> >
> > Topics of the day:
> >
> > 1. AW: DC-ENVIRONMENT Digest - 30 Aug 2002 to 2 Sep 2002
> (#2002-18)
> >
> >
> ----------------------------------------------------------------------
> >
> > Date: Mon, 9 Sep 2002 13:39:24 +0200
> > From: Thom Pick <[log in to unmask]>
> > Subject: AW: DC-ENVIRONMENT Digest - 30 Aug 2002 to 2 Sep
> > 2002 (#2002-18)
> >
> > thanx simon,
> >
> > i understand that we should look into the description of value =
> > structure if
> > we discuss observations and measurements in DC-ENV context.
> however, i
> > wanted to take it one step further down by asking whether
> we need to =
> > include
> > description on observation and measurements for environmental =
> > information
> > resources in a DC-ENV application profile AT ALL. what kind of =
> > resources
> > would that be, if not structured data that are dealt with by expert
> > communities, like e.g. OGC in there own ways.=20
> >
> > most reports on the state of the environment are based on
> structured =
> > data.
> > but the data that make it into the report are mostly heavily
> > aggregated =
> > and
> > 'refined'. since the report usually refers to the source data
> > it may be
> > useful to use DC to describe them as well (in addition to the
> > report) =
> > if
> > they are available online. but as of now i'm not quite sure about =
> > example
> > cases.
> >
> > so if anybody on the list could give us some hint here....!
> >
> > cheers
> >
> > thom
> > -----Urspr=FCngliche Nachricht-----
> > Von: Simon Cox [mailto:[log in to unmask]]
> > Gesendet: Dienstag, 3. September 2002 03:01
> > An: [log in to unmask]
> > Betreff: Re: DC-ENVIRONMENT Digest - 30 Aug 2002 to 2 Sep 2002
> > (#2002-18)
> >
> >
> > No - I'm not proposing that simply importing the namespace is the =
> > solution.
> >
> > Just pointing out that there has been some recent work in
> this area, =
> > and the
> > conceptual models are now reasonably well understood.
> > In particular, there are some solid ideas about how to structure a
> > description of complex observations and measurements.
> > This includes a description of the value structure (the "column =
> > headings")
> > which might be just what you are looking for.
> > In return, the structured-data community could probably gain some =
> > valuable
> > assistance in the semantic aspects of the problem.
> >
> > Simon
> > _____
> > [This mail represents part of a discussion of work in progress
> > and should not be used for any purpose without my permission.]
> > _____
> > [log in to unmask] CSIRO Exploration & Mining
> > 26 Dick Perry Avenue, Kensington WA 6151
> > PO Box 1130, Bentley WA 6102 AUSTRALIA
> > T: +61 (8) 6436 8639 F: +61 (8) 6436 8555 C: +61 (4) 0330 2672
> > http://www.csiro.au/page.asp?type=3Dresume&id=3DCoxSimon
> >
> > > -----Original Message-----
> > > From: Automatic digest processor [mailto:[log in to unmask]]
> > > Sent: Tuesday, 3 September 2002 7:47 AM
> > > To: Recipients of DC-ENVIRONMENT digests
> > > Subject: DC-ENVIRONMENT Digest - 30 Aug 2002 to 2 Sep 2002
> > (#2002-18)
> > >
> > >
> > > There is one message totalling 185 lines in this issue.
> > >
> > > Topics of the day:
> > >
> > > 1. AW: DC-ENVIRONMENT Digest - 9 Aug 2002 to 29 Aug 2002
> > (#2002-16)
> > >
> > > =
> >
> ----------------------------------------------------------------------
> > >
> > > Date: Mon, 2 Sep 2002 10:26:38 +0200
> > > From: Thom Pick <[log in to unmask]>
> > > Subject: AW: DC-ENVIRONMENT Digest - 9 Aug 2002 to 29 Aug
> > > 2002 (#2002-16)
> > >
> > > simon,
> > >
> > > i reckon that you're proposing to point to the OWS namespace
> > > in order =3D
> > > to
> > > include measurement data.=3D20
> > >
> > > however, my intention was to start a discussion on whether the =3D
> > > 'element'
> > > measurement data is needed at all for the discovery of
> environmental
> > > information...... keeping in mind that we are discussint (the
> > > potential =3D
> > > need
> > > of) a dc-env application profile. it seems to me that we
> basically =
> > =3D
> > > could
> > > make do with DCES and DCQ if we agreed on some additional
> > > schemes for =3D
> > > the
> > > environmental domain.
> > >
> > > thom
> > >
> > > -----Urspr=3DFCngliche Nachricht-----
> > > Von: Simon Cox [mailto:[log in to unmask]]
> > > Gesendet: Freitag, 30. August 2002 05:20
> > > An: [log in to unmask]
> > > Betreff: Re: DC-ENVIRONMENT Digest - 9 Aug 2002 to 29 Aug 2002
> > > (#2002-16)
> > >
> > >
> > > Concerning your question
> > >
> > > > 1. what is your opinion on the element 'Measurement Data'?
> > > >
> > > > Name: measurementdata
> > > > Label: MeasurementData
> > > > Definition: Describes the content, function and format
> of a set of
> > > > measurement data
> > > > Comment: May be helpful to integrate packages of
> measurement data.
> > > >
> > > > however, it seems to me that this could be covered by DCMES
> > > > appropriately.
> > >
> > > I am working with Open GIS Consortium on a very closely related =
> > topic.
> > > I am responsible for a discussion paper on this topic
> > > published by OGC =3D
> > > at
> > >
> > > http://www.opengis.org/techno/discussions/02-027.pdf
> > >
> > > The paper was developed as part od the SensorWeb Enablement
> > > thread of =3D
> > > OGC's
> > > Web Services interoperablity project, which included US EPA,
> > > NASA, NIMA
> > > amongst its sponsors. The paper is currently under revision
> > > as part of =3D
> > > a
> > > follow-up project in OGC. A focus of the revision is to
> harmonise =
> > =3D
> > > better
> > > with some external standards in this area, such as [1] and
> > also with =
> > =3D
> > > OGC's
> > > Web Coverage Service specification which deals mostly with
> > > gridded data =3D
> > > and
> > > imagery. There may be some adjustments to terminology in the =3D
> > > discussion
> > > paper as this harmonisation proceeds. Note that an important
> > > part of =3D
> > > the
> > > "metadata" of an image or coverage is a description of the
> > > "range set", =3D
> > > in
> > > both semantic and structural terms.
> > >
> > > You would probably also want to look at ESML from NASA.
> > >
> > > Anyway, my suggestion would be, rather than inventing
> yetanother =3D
> > > system, you
> > > should adopt or adapt something from this area.
> > >
> > >
> > > [1] National Research Council. 1995. "Expanding the
> Vision of Sensor
> > > Materials".
> > > Committee on New Sensor Technologies: Materials and
> > Applications. =3D
> > > National
> > > Academy Press. Available [online]:
> > > <http://books.nap.edu/books/0309051754/html/index.html>
> > >
> > >
> > > > -----Original Message-----
> > > > From: Automatic digest processor
> [mailto:[log in to unmask]]
> > > > Sent: Friday, 30 August 2002 11:07 AM
> > > > To: Recipients of DC-ENVIRONMENT digests
> > > > Subject: DC-ENVIRONMENT Digest - 9 Aug 2002 to 29 Aug 2002
> > > (#2002-16)
> > > >
> > > >
> > > > There is one message totalling 69 lines in this issue.
> > > >
> > > > Topics of the day:
> > > >
> > > > 1. dc-environment application profile
> > > >
> > > > =3D
> > > =
> >
> ----------------------------------------------------------------------
> > > >
> > > > Date: Thu, 29 Aug 2002 14:22:44 +0100
> > > > From: Thom Pick <[log in to unmask]>
> > > > Subject: dc-environment application profile
> > > >
> > > > dear colleagues,
> > > >
> > > > stefan strathmann of the goettingen state and university
> > > > library has been
> > > > looking into formulating an application profile for the
> > > environmental
> > > > domain.
> > > >
> > > > the essence of the study is that the DC-Environment
> > > > Application Profile
> > > > would consist of the following namespaces:
> > > >
> > > > =3D3DB7 Dublin Core Metadata Element Set, Version 1.1
> > > > [http://purl.org/dc/elements/1.1/]
> > > > =3D3DB7 Dublin Core Qualifiers [http://purl.org/dc/terms/]
> > > > =3D3DB7 DC-Library Metadata Element Set (DC-LMES)
> > > > =3D3DB7 DC-Library Metadata Element Set Qualifiers (DC-LMES
> > > Qualifiers)
> > > > =3D3DB7 DC-Environment Metadata Element Set (DC-EMES)
> > > > =3D3DB7 DC-Environment Metadata Element Set Qualifiers (DC-EMES
> > > > Qualifiers)
> > > >
> > > > his proposal for the DC-EMES includes the new element
> > > > 'Measurement Data'
> > > > and recommends to agree on schemes for the element
> Subject (e.g. =
> > the
> > > > General Multilingual Environmental Thesaurus, GEMET),
> Relation and
> > > > Coverage.
> > > >
> > > > I would like to request your input on the following questions:
> > > >
> > > > 1. what is your opinion on the element 'Measurement Data'?
> > > >
> > > > Name: measurementdata
> > > > Label: MeasurementData
> > > > Definition: Describes the content, function and format
> of a set of
> > > > measurement data
> > > > Comment: May be helpful to integrate packages of
> measurement data.
> > > >
> > > > however, it seems to me that this could be covered by DCMES
> > > > appropriately.
> > > >
> > > > 2. what is your opinion on using GEMET (or its monolingual
> > > > equivalent) as
> > > > a scheme for the element Subject?
> > > >
> > > > yes, i know it is BIG (> 5000 terms) but it integrates
> > > terms from all
> > > > fields of the environmental domain...
> > > >
> > > > 3. which schemes would you propose for the element qualifiers
> > > > Coverage:Spatial and Coverage:Temporal?
> > > >
> > > > will DCMI Box and DCMI Period suffice? will we need a 'Geo =3D
> > > Thesaurus'?
> > > >
> > > > 4. should the elements Audience and Availability be included?
> > > >
> > > > we intend to discuss these issues also at the dc-env
> > interest group
> > > > meeting at dc-2002. since we don't have much time available
> > > > there and then
> > > > it would be great if we could get the discussion started on
> > > > the list now.
> > > >
> > > > cheers
> > > >
> > > > thom
> > > >
> > > >
> > > > ps: i have uploaded the draft version to the interest group
> > > > file server.
> > > > it is available at
> > > > http://www.jiscmail.ac.uk/files/DC-ENVIRONMENT/DC-Env_AP_First
> > > _Draft.doc
> > >
> > > -----------------------------
> > >
> > > End of DC-ENVIRONMENT Digest - 9 Aug 2002 to 29 Aug 2002
> (#2002-16)
> > >
> *******************************************************************
> > >
> > > ------------------------------
> > >
> > > End of DC-ENVIRONMENT Digest - 30 Aug 2002 to 2 Sep 2002
> (#2002-18)
> > >
> *******************************************************************
> > >
> >
> > ------------------------------
> >
> > End of DC-ENVIRONMENT Digest - 3 Sep 2002 to 9 Sep 2002 (#2002-20)
> > ******************************************************************
> >
>
|
