[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]
Subject: Re: [egov] Metadata-template for taxonomies
The whole European Parliament's work on object
identification, registration and metadata management is using ebXML CCTS and ISO
11179....I would indeed strongly recommend it.
That doesn't mean that there aren't a whole load of
business-level implementation issues to be solved, which current OASIS TCs do
not address.
I am drafting a scoing paper as agreed at the
Tuesday conf-call, but I offer to the list the simple "use case" I mentioned at
the meeting:
- we have dispersed and un-coordinated projects
needing to identify objects (documents, projects, people, applications, you name
it....if you excuse the pun...)
- some want/need to implement the SGML "Formal
Public Identifier" (FPI); others the DOI or other Handle systems; others still
are looking at XRI, TopicMaps PSIs, etc...
- everyone wants metadata, standradised,
structured, validated, predictable...
- many have their "instant solution" up the sleeve
and ready to seduce the unwary;
- we cry "standards, standards, this way..." but
they don't come or don't understand;
- we want business decisions on technology before
IT decisions on the business;
- none of the standards "show" you how or why you
should be interetsed/worried/scared to death about: naming conventions; object class definitions; criteria for creating
taxonomies/classifications; best practices for configuring/customising possible
C.O.T.S. to conform with business policies.
- CCTS and the ebXML methodology guides you
through process-driven discovery, encapsulation and registration of core
components, BIEs, CPPs, etc. All great stuff, and that will indeed help us to
propose solutions but the *technical specifications* alone do not help us, even
if they are, with UBL, BCM, etc, making great efforts to explain the "why" of
the specs. We need also to:
- get recognition of these approaches on the desks
of senior management;
- get those managers engaged in the semantics (and,
frankly, get the technicians off, for the moment);
- explain *why* (rather than how), in *business*
rather than technical terms: explain the "how" in human/business language, not
UML, UMM, etc.
- offer quantifiable CBAs demonstrating the
business value and ROI;
To put it succinctly (philosophers need not
reply); we need a "how of the why", not a "why of the how".
Peter
Peter Brown
(Head of Information Resources Management European Parliament) ______________________ This mail reflects the views and opinions of the
author alone. Institutional affiliation is indicated for information purposes
only. Any formal correspondence on any matter contained within this mail should
be addressed to pbrown@eurpoarl.eu.int
----- Original Message -----
From: CRAWFORD, Mark
Sent: Friday, January 30, 2004 9:30 PM
Subject: Re: [egov] Metadata-template for taxonomies Mark Crawford Research Fellow - LMI XML Lead W3C Advisory Committee, OASIS, RosettaNet Representative Vice Chair - OASIS UBL TC & Chair Naming and Design Rules Subcommittee Chair - UN/CEFACT XML Syntax Working Group Editor - UN/CEFACT Core Components ______ Logistics Management Institute 2000 Corporate Ridge, McLean, VA 22102-7805 (703) 917-7177 Fax (703) 917-7481 Wireless (703) 655-4810 mcrawford@lmi.org http://www.lmi.org "Opportunity is what you make of it" -----Original Message----- From: Carl Reed <creed@opengis.org> To: Duane Nickull <dnickull@adobe.com>; Maewyn.Cumming@e-Envoy.gsi.gov.uk <Maewyn.Cumming@e-Envoy.gsi.gov.uk> CC: egov@lists.oasis-open.org <egov@lists.oasis-open.org>; Michael Bang Kjeldgaard <mbk@itst.dk> Sent: Fri Jan 30 11:33:15 2004 Subject: Re: [egov] Metadata-template for taxonomies All - Just to throw one other element into this discussion. The ISO TC 211 (geospatial) and the Open GIS Consortium both use the ISO 19115 (metadata for things spatial) standard as the abstract model foundation for expressing spatial metadata. As many of the current and emerging E-Government applications have a spatial component, perhaps 19115 should also be considered (Please see note below). Carl Reed OGC ISO 19115 - Metadata, is a formal schema for geospatial metadata that is intended to apply to all types of spatial information. ISO 19115 provides a UML model of metadata, based on the US Federal Geographic Data Committee's (FGDC's FGDC) Content Standard. Its chief purpose is to support profiles, using a small set of required elements and many optional ones. (ISO 19115 is largely harmonized with the Dublin Core and in the US, the FGDC Content Standard for Digital Geospatial Metadata.) Many government organizations have developed data and metadata that conform to these standards. All future registered ISO/ TC211 metadata profiles must include these core elements, to ensure interoperability and to guarantee productive searches. For the purpose of spatial catalogs, the schema and core elements of ISO 19115 must be implemented by conforming implementations. A future metadata implementation schema (ISO 19139) in progress and will soon offer implementation guidance. ----- Original Message ----- From: "Duane Nickull" <dnickull@adobe.com> To: <Maewyn.Cumming@e-Envoy.gsi.gov.uk> Cc: <egov@lists.oasis-open.org>; "Michael Bang Kjeldgaard" <mbk@itst.dk> Sent: Wednesday, January 28, 2004 11:21 AM Subject: Re: [egov] Metadata-template for taxonomies > Maewyn: > > The beauty of the ebXML Registry is you can use multiple metadata > standards to classify and associate objects. Dublin core, IMO, does not > align perfectly with the metadata requirements of either ISO/IEC 11179 > 2002 part 3 or the ebXML RIM however. In fact, there are several > examples of it in your email. ISO/IEC 11179 (in part 3) specifies a set > of metadata for each object referencable form a MDR. The attributes for > coverage are user defined and extensible while DC supports geospatial > and temporal only. Another example is the lack of a status attribute > (mandatory in 11179, not present in DC). > > I would suspect if DC is a useful standard for classifying registry > objects among other such standards. I see no reason why it also cannot > be used alongside other MD standards for taxonomies, ontologies etc. > > Duane > > Maewyn.Cumming@e-Envoy.gsi.gov.uk wrote: > > > > > I would like to see us sticking to Dublin core (ISO 11179) as far as > > possible for this. Comments inline: > > > > > > Before we finalise this it might be useful to decide what we will be > > using the metadata for: we want something slightly different if its to > > be used for searching (this means building a search interface that > > allows field searching) or if its going to be there as background > > information that is fairly visible to users. If its to be for > > searching we need to create some controlled lists (a taxonomy for > > taxonomies!) so that it actually works. We might also want some > > subject terms. > > > > > > > > > > Duane Nickull <dnickull@adobe.com> > > > > 27/01/2004 18:00 > > > > To > > Michael Bang Kjeldgaard <mbk@itst.dk> > > cc > > egov@lists.oasis-open.org > > Subject > > Re: [egov] Metadata-template for taxonomies > > > > > > > > > > > > > > > > > > Michael: > > > > This model can all be declared using existing ebXML RegRep or ISO > > 11179 metadata. I will comment inline: > > > > Michael Bang Kjeldgaard wrote: > > > > > Title (local title) > > > > > RIM has a name attribute for this. > > MC: DC specifies 'Title' > > > > > Title (English translation + plus > > explanation if needed) > > > > > RIM supports multilingual titles, names and descriptions. > > MC: DC calls for 'Alternative title' > > > > > Coverage (domain / area of use > > covered by the taxonomy) > > MC: DC Coverage' is specifically for geospatial or temporal > > coverage, so would be appropriate for area of use > > > > > Taxonomy and the objects relationships to other objects within a > > specific or multiple taxonomies can be declared or asserted via > > associations and/or classifications in the ebXML RIM. > > > > > Description: (e.g. how and how > > widely it is used, how old it >is, comments on quality) > > > > > Description is a mandatory attribute for each registry object. The > > "how old it is" is expressed as part of the audit trail. The audit > > trail is a better way of declaring this information because it may be > > necessary > > to determine how old an object is in its' current or a previous > > status. Registry objects can have different status attribute values > > throughout their life cycle (Submitted, Approved, Superseded, etc...) > > MC: DC: suggest using Date.issued to clarify how old a taxonomy is. > > DC doesn't have a 'status' element, we use it in the UK govt though as > > it is quite valuable. It would be valuable to have a stated list of > > values for this element > > These two elements could then be incorporated into an audit trail. > > > > > Type: (taxonomy, ontology, glossary > > etc.) > > > > > RIM attribute ObjectType. This is user extendable to account for any > > anomolies users may wish to express. > > > > MC: DC calls for 'type'. Again, it would be valuable to have a stated > > list of values for this element. I will be sending a list of > > definitions for types of controlled vocabularies by a seperate e-mail, > > they are the definitions used by the British Standard for thesauri. > > > > > > > Volatility (how often does it get > > updated/changed) > > > > > This is handled via the Audit trail. Each item has a complete > > auditable event log associated with it. > > > > MC: DC has nothing for this. In UK gov we use Date.updating > > frequency, which is more specific than 'volatility' > > > > > Editor (responsible organization / > > contact person) > > > > > Called "Owner". There is both an organization and actors associated > > with that organization. > > MC: DC calls for 'Creator' > > > > > Government use (international?, > > nation?, government tier or >domain?) > > > > > Can be expressed via the classification schemes or its' relationship > > to other items. For example - you could state that a specific data > > element is used within a schema by a certain government department. > > MC: DC Seems to repeat some of what's covered under 'Coverage' > > > > > Audience (targetgroup: > > IT-professionals, public sector employees, business, citizens) > > > > MC: I always advise caution when using 'audience'. To use it for > > finding taxonomies you need a controlled list of audiences, also most > > taxonomies are really designed for specific audiences, so this tends > > to tune into a meaningless jumble. > > > > > > Links (to the sourcefile or a > > webservice + to implementations, e.g. online portals, specific > > services etc.)handled via associations. > > MC: DC uses 'Relation' with specific types of relation e.g. > > IsVersionOf, Is FormatOf > > > > ebXML RIM can do this entire metadata with ease and more I suggest > > that we use the ebXML RIM within this group since it appears to meet > > all our requirements plus it is extensible in a forwards compatible > > manner. > > > > Duane Nickull > > > > -- > > Senior Standards Strategist > > Adobe Systems, Inc. > > http://www.adobe.com > > > > > > > > > > To unsubscribe from this mailing list (and be removed from the roster > > of the OASIS TC), go to > > http://www.oasis-open.org/apps/org/workgroup/egov/members/leave_workgroup.php. > > > > > > PLEASE NOTE: THE ABOVE MESSAGE WAS RECEIVED FROM THE INTERNET. > > > > On entering the GSI, this email was scanned for viruses by the > > Government Secure Intranet (GSI) virus scanning service supplied > > exclusively by Energis Communications in partnership with MessageLabs. > > > > GSI users see > > http://www.gsi.gov.uk/main/notices/information/gsi-003-2002.pdf for > > further details. In case of problems, please call your organisational > > IT helpdesk. > > > > -- > Senior Standards Strategist > Adobe Systems, Inc. > http://www.adobe.com > > > > > To unsubscribe from this mailing list (and be removed from the roster of the OASIS TC), go to http://www.oasis-open.org/apps/org/workgroup/egov/members/leave_workgroup.php. > To unsubscribe from this mailing list (and be removed from the roster of the OASIS TC), go to http://www.oasis-open.org/apps/org/workgroup/egov/members/leave_workgroup.php. To unsubscribe from this mailing list (and be removed from the roster of the OASIS TC), go to http://www.oasis-open.org/apps/org/workgroup/egov/members/leave_workgroup.php. |
[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]