I have a number of comments about the first steps/draft.
Sorry about the length! </txk>
1. Do we have agreement about the top-level of the
heirarchy? What is missing? What should not be
there?
Suggestions for top-level categories:
DOCUMENT
IMAGE
SOUND
SOFTWARE
DATASET
GAME (new)
TALK (new)
VR (new)
I have not included MISC. as per my comments
below.
Someone else recommended GAME since this
is and will continue to be a growing area of
Web activity in the future. More below. Yes? No?
TALK are web talk pages or pages which are
entry points into a web-based "conversation".
Yes? No?
I have added VR as a top-level category here
since VR is certainly more than just images
but are more like environments complete with
sounds and documents and images. IS there a
need for a top-level like this?
2. I think we should try to work through a general
outline of the top two levels of the heirarchy first,
making reference to the lower levels as we go. But
if we munch all of this together, I think we run the
risk of letting the granularity issues overwhelm us.
3. Extensibility: Use Scheme qualifier to specify the use
of thesauri. DC Resource Types could be the default scheme.
But I don't think we should be grabbing too many terms
from specialized knowledge domains if we can avoid it.
Jon's suggestion about local modifications using x- is
useful. But I think the purpose of having DC resource types
would be limited if we don't try to impose a general
structure for the top 2, maybe three levels.
4. I have some qualms about the top category "Miscellaneous"
since this quickly becomes a garbage can category. I
don't think that it is useful as a default type. I would
rather there be no default be "none" or freetext or even some
"document.unknown" type than "misc." I used to use MISC
in organizing my own files and I think that it breeds sloppiness.
5. I agree with some of the other comments about folding
some of the categories into one. We should work to remove
ambiguous classes.
Examples:
document.article.essay
document.article.techreport
document.article.story
I don't know if the distinction "referred/unreferred"
is a separate class. If we do "document.article.referred"
then referred is the same as "essay" or "techreport".
This would not make sense. I would rather drop it for now
since I think the distinction refers to the process of
attributing value and authority to the publication. If we
do this then why not "document.article.edited" or
"document.article.
Also:
"document.monograph.thesis" rather than "document.thesis"?
I don't like serial as it describes a quality of publication
rather than a class of the content.
For example:
document.article.serial
document.monograph.serial
document.poem.serial
document.press-release.serial
The subcategories here (journal, magazine, newsletter,
newspaper, proceedings) all work as classes of documents
like "document.newspaper" but is this really what is
going to be described? Or is it more likely that
a discrete part of each will be described, i.e. the
article rather than the journal? These are classic
"paper" classes but I wonder if they work to describe
a discrete resource very well?
If we need a category to describe each, then it should be:
document.newspaper
document.proceedings
document.journal
document.newsletter
There is something that I don't like about this approach tho'...
6. Missing in the current draft are some notes about what is distinctive
about these categories. Hopefully they are self-evident but if not, then
they should be described. And again, we should try to use a descriptive
note to indicate where people can fold things into a category.
For example:
"DOCUMENT.ARTICLE
Objects where the contents are of such length
and/or self-contained in subject matter that
the author would consider this as an article.
The purpose of the contents is to fully realize
a particular objective in a relatively concise
form. Also implied is that the document exists
within the context of a collection of other
articles which are made available serially.
This class includes essays, stories, preprints,
and other short written forms."
7. There are some categories that I think are
important (just from a random survey of websites) and
which are missing from the list. I think they are
important because I see alot of webpages out there
that would be usefully described by the below:
DOCUMENT.PROMOTION
Primary purpose of document is to promote or provide
information about a product or service, i.e. advertising.
DOCUMENT.MEDIA
Primary purpose of document is a press release or information
for media consumption.
DOCUMENT.BIBLIOGRAPHY
Document consisting of an organized list of
references to other resources, either in
print or electronic form.
DOCUMENT.INDEX
Document intended to provide an index to services
or documents within a website, or some subset of
documents.
DOCUMENT.DICTIONARY
Document consisting of an organized list of terms
and their definitions.
DOCUMENT.ABSTRACT
A summarization of another document.
DOCUMENT.PAMPHLET
A short document intended to provide general
information on a single topic or subject, i.e.
a pamphlet, booklet or factsheet of the sort that
public agencies or companies give out.
DOCUMENT.BIOGRAPHY
Primary purpose of the document is to provide information
about a person or persons.
DOCUMENT.HELPFILE
Includes manuals or other pages intended to provide assistance for
software/hardware.
DOCUMENT.NAVIGATION
The primary purpose is to provide a gateway or form
interface to local or remote navigation services.
DOCUMENT.FORM
A form of some type, i.e. comments, survey, order requiring
input of a user.
DOCUMENT.SCREED (Sorry Peter! I still think it is a useful
catch-all for lots of the "personal" webpage contents).
An informal piece of writing; fragment or other
type of personal discourse.
8. I think the IMAGE category is generally good but still
believe that the following should be there:
IMAGE.MAP
(Representation of cartographic informtion)
IMAGE.SCORE
(Representation of a musical score)
IMAGE.ENGINEERING
(Representations of engineering diagrams, i.e. CAD-CAM)
IMAGE.CHART
(Representations of various forms of charts,
i.e. gantt, org charts, flowcharts)
Also, SOUND should also have sound.broadcast since
I know that alot of radio stations are doing
broadcast on the net (which may have speech/music/
ambient sounds).
I think that SPEECH is better than NARRATION since
SPEECH is any expression of thought in spoken words
whereas NARRATION is a kind of SPEECH. I am thinking
of SPEECH as an utterance rather than as an address.
9. SOFTWARE
Although game is generally useful, I think that
games are going to be a significant network business
and will web-based as well. This is why I suggest
GAME as a top-level category. For example:
GAME.SOFTWARE
GAME.SOFTWARE.DEMO
GAME.NETWORK
Yea? Nay?
10. Webpage: everything is a webpage. I think
homepage is the better and common description of a
top-level page to some organization or person's website.
In any event, it needs a note to qualify the scope of the
resource-type.
webpage.organization = document.homepage.organization
webpage.personal = document.homepage.personal
homepage.personal is not the same as document.biography
Enough for now...
</txk>
---------------------------------------------------------------
Mr. Terry Kuny Home Office: 819-776-6602
XIST Inc./ Email: [log in to unmask]
Global Village Research URL: http://xist.com/kuny/
Snail: Box 1141, St. B, Hull, Quebec, Canada
---------------------------------------------------------------
|