I'm glad to find such project which is designed to express the subject
content of Web resources in terms of the interrelationship of subject
terms. Exactly with the same objective, I've worked on developing an
extended schema [3] of SKOS (Simple Knowledge Organization System) to
express the relationships which exist between two subject terms.
Definitely "isAbout" is very generic way of defining the subject
matter of the resources. In most cases there will be more than one
subject descriptors (keywords) for a resource. These descriptors can
be joined with one another with certain "relations". In Ref. 1 five
basic relations (facets) has been identified such as "Discipline",
"Entity", "Property", "Action" and Modifiers.
The subject matter is expressed in natural language in any document
(resource). "The meaning of some information can be presented in a
number of ways i.e. in the form of many sentences. If we consider the
various sentences for the same information (surface structure), there
should be a unique deep structure for the various surface
structures" (Prasad, 1993). This unique deep structure constitutes of
multiple facets. Ref 2 and ref. 3 are based on the work done by Prof.
Ganesh Bhattacharyya on Subject Indexing Languages Ref.1.
With a brief introduction on my work, I would like to join the
project. I look forward for comments on my post. Please refer to the
references (some of which I'm attaching in the project site).
Reference:
[1] G. Bhattacharyya. (1981). Subject Indexing Language: its theory
and practice. In DRTC Refresher Seminar - 13, New Developments in LIS
in India, 14-17 October 1981, DRTC, Bangalore, India, 1981.
[2] A.R.D. Prasad. (1993). Application of computer-based natural
language processing tools and techniques in developing subject
indexing languages. Thesis submitted to the Karnatak University,
Dharwad for the award of Doctor of Philosophy in Library and
Information Science. Karnatak University, Dharwad. pp. 40.
[3] Nabonita Guha & A.R.D. Prasad. (2007). Expressing Faceted Subject
Indexing in SKOS/RDF. In International Conference on Semantic Web &
Digital Libraries (ICSD-2007), 21-23 February 2007, Bangalore, India,
2007.
On Jul 16, 3:49 am, "guha.nabon...@gmail.com"
Nabonita,
Excellent! And welcome to the project. We are very fortunate to have
your expertise.
I will review these articles and directly comment soon.
We will also begin some substantive topics in this forum this week. I
look forward to your comments and participation.
Mike
Hi Nabonita,
I have read your papers with great interest. Thank you for posting
them so others can easily find them.
I found the Bhattacharyya reference (Ref 1) to be a helpful
background. I have been somewhat familiar with Colon Classification
and Ranganathan's work (I've cited in some older papers), but I was
not familiar with POPSI and the Classaurus. There often times seems
to be too much distance between those in information architecture and
library science with those in computer science more traditionally
defined.
This reference was a good basis for then reading your own Ref 3; very
helpful and very interesting. I have some specific questions (in no
particular order):
1. What has been the reaction of the SKOS community (esp Miles and
Brickley) to the proposed language extensions?
2. How do you see your proposed approach relating to the
"unclassified" 'core' subjects within the proposed UMBEL, and then to
its 'unofficial' classified, hierarchical version?
3. What reference sources can you propose could contribute to the
standard listing of "bag of subjects" proposed for the core UMBEL (see
other posting)?
4. What differences do you see from your specific SKOS proposal and
UMBEL as you understand it (I'm thinking particularly about inclusion
or not of Actions)?
Thanks, Mike