Re: Extension of SKOS

Hi Nabonita!

you wrote:

> As I was studying SKOS, I felt that there must be some mechanism in SKOS
> by which the relationships between various subject components of a
> documents can be represented. 

Can you please give an example? If I understand you right then you are
talking about syntactic indexing. I am just searching for literature
about the Universal Decimal Classification (UDC) - a classification that
allows you to build numbers that indicate the relationships between
concepts of a topic.

For instance if you have a book on the 19th Century History of the U.S.
then its number is:


94"18"(73)

94         History
"18"       19th century (Common auxiliaries of time)
(73)       United States (Common auxiliaries of place)


Or about the organization of Arabian Islam in the United States the UDC
number should be:

28-7(=411.21)(73)

It's build of:

28         Islam
-7         Religious organization and administration (special Table)
(=411.21)  Arabs (Common auxiliaries of ethnic grouping and nationality)
(73)       United States (Common auxiliaries of place)


Or even more complex: Research about Education and Music and their
relationship to Mental developement:

[37+78]:159.92

It's build of:

37       Education
+        Combination
78       Music
[...]    Grouping
159.92   Mental development and capacity. Comparative psychology

See more about UDC at http://www.udcc.org/outline/outline.htm

Sadly UDC is not freely accesible on the web so you probably have to
visit your library and get a book - this may be scary for the computer
scientists that a working on the semantic web not having heard of UDC
before ;-) By the way UDC is used for about 100 years!

> RDF has the potentials to do so apart from standardising the
> vocabulary which is equally inportant. With the combination of
> both controlled vocabulary and faceted representation the
> true purpose of Knowledge Organization can be achieved.
> 
> We can start such initiative by which the facetized approach of library
> classification can be also incorporated in SKOS. I would like to know
> the opinion of you all regarding this issue.

We should definitely find a way to encode UDC in expanded SKOS. This
will also include general methods of coordination - a feature that is
needed anyway.

Greetings,
Jakob

Received on Monday, 31 July 2006 14:23:34 UTC