RE: Re: Defining basic vocabulary of faceted classification (TERMINOLOGY)

Luca,
so far this discussion is all over the place touching very
different levels and issues.
As for the facets, arrays(subfacets) and foci, at least for what is in
the library science literature they are very technical and
conceptual terms - I don't think they serve the right purpose
when detached from the classificationist's jargon.

The explanation facet analysis (in laymans' terms)if applied
in building metadata standard/structure.

Every metadata is an object surrogate that serves some
management or discovery function. Therefore there will be need to describe
a) object itself (its intellectual content, authoriship/provenance, and
physical characteristics)

b) technical requirements needed to use or manipulate the object
(e.g.if it is information object stored as an image file ....one would
probably need to
state software, hardware requirement - this metadata is usually inserted and
read by
machines

c) meta-metadata or administrative metadata - some information on the
metadata production
who did description, when, which agent and service...this data is only for
the purpose of
metadata managment

Then if you focus on an OBJECT only. You will have number of attributes
pending to be expressed
depending what the metadata standard is about.
I will use the example of the information object as I am familiar with these
metadata
standards.
If it is an information object you will have to describe, for instance, at
least
- object's content,
- object physical characteristics (instantiation)
- object intellectual property rights.

These three groups/classes of attributes don't overlap as they serve
different purpose in resource discovery.
Let's take for instance object content... this content has some title, it
has some subject,
it has some presentation form, audience, purpose. So you will define a
number of descriptive
elements to cover this with clear mutually exclusive semantic although they
all belong to "content
group" of elements as this one have clearly defined function in information
exchange and management.

The next group/class of descriptive elements are elements that describe
instantiation of this content
(date,format,identifier)
Your group of intellectual property rights would state intellectual
responsibility for this object,
name of the creator(s), contributor(s), publishers, rights and condition of
use
This is simplified analysis based on Functional requirement for
bibliographic data (http://www.ifla.org/VII/s13/frbr/frbr1.htm), where each
information object (document)
is said to consist of work, expression, manifestation, item. Many metadata
standards
for network resource discovery are built based on this or similar approach
to modelling (e.g. Dublin Core,
Learning Object Metadata, INDECS etc.)

Within any of these group of elements you can have further structuring...
Let's take the simpliest example. E.g. creator... which can be a person or
corporate body. The creator can
have role (some kind of authoriship), if it is individual it can have
name/surname/afiliation/address/e-mail
(e.g.VCard).
Or e.g. title, which can be original,translated, full,
abbreviated,alternative, hidden. Then
the title can have one of more subtitles etc...
Within "content group" you can have one or several elements describing
subject i.e. what the object is about
The subject can be a person, a place but can also be abstract or concrete
concept of some kind. Very often subject is complex and composed of
different facets: the object is about som
kind of entity on which some action is performed by certain agents with
certain products.
The value in subject elements can be taken from classification schemes,
thesauri, subject-heading system etc.
This means that the value in the subject field is not taken from the object
itself but is 'imported' from
another system. To add the confusion some call this specific part of
metadata 'subject metadata'
...yet another meta level (?!).

In somma...

Metadata standards tend to have structure and organization of elements that
reflects their purpose. They
are very differently structured. This structure can be more or less
hierarchical. In the broader meaning
of the "facet analysis" it could be said that metadata standard can have
'faceted structure'.
Hence the need to handle facets on the several levels of metadata.

I hope I managed to clarify....but it could be that I am looking into the
issue of XMFL from entirely
wrong angle which makes it hard to understand. I would appreciate any
suggestion (any useful article...?)

Aida

>> 1)FIRST application - general metadata structure (classes of descriptive
>> elements supporting
>> a particular function)... semantic of these elements have to be mutualy
>> exclusive e.g. every metadata standard will have elements supporting
>> discovery, administration
>> and technical features of the object. Elements supporting e.g. discovery
>> will be grouped into broad
>> classes of elements that describe e.g. content, intellectual property and
>> instantiation.

>the mutually exclusive principle is about only the "elements level"
>resulting from main division (discovery, administration etc.) or also the
>inner "broad classes" (content etc.)?




----------
Thanks for playing. To unsubscribe from this group, send an email to:
[email protected]



Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/





----------
Thanks for playing. To unsubscribe from this group, send an email to:
[email protected]

 

Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/