Proposed agenda for the "Taxonomy Mgt and Support for Semantic
Search" Telecon
Max: The background section is intended to identify
the rationale for my recommendation to focus Monday's meeting on the topic
of OWL and its relationship with our Taxonomy Mgt Requirements. Could you
please review and confirm that you agree with the proposed course of
action.
Massimo/Katia: Could you please look at this note as
I'm suggesting that we take up your offer for a presentation. This is of
course dependent on Max's agreement with the proposed course of
action.
Call Details
Mon 19 Jan at 13:00 Pacific.
Toll free: +1 877-272-5572
International: +1 866-453-5550
Participant code: 497-1748
UTC: Mon-21:00, Seattle:
Mon-13:00, New York: Mon-16:00, London: Mon-21:00,
Frankfurt: Mon-22:00, Moscow: Mon-24:00, Tokyo:
Tue-06:00, Sydney: Tue-08:00, Auckland:
Tue-10:00
Background
When we started out on Requirements
011-014 [1] it was with the intent to deal with UDDI's taxonomy management
needs; line items 1-11 of [1] identified a fairly crisp set of requirements
which we prioritized. We have since embarked in looking at Semantic Search
capabilities which, given OWL's relevance to our taxonomy needs, naturally
led us at looking at the possibility of providing semantic search
capabilities. That said, we never prioritized semantic search requirements
and need to do so.
I think that looking at Semantic
Search capabilities is worth it, though we need to keep our sights on the
initial set of requirements which were to (in short): define a means of
publishing taxonomies; query taxonomies; manage taxonomies; and define a
normative schema for taxonomies. Solutions need to be in the form of APIs
and schemas which satisfy the requirements we identified and prioritized in
[1]. From my research, OWL does seem to fit the bill and is clearly timely
although I'm concerned whether it's premature for us to leverage it for a
number of reasons (stability of the spec; willingness of vendors to
implement it; etc. - keep in mind that in order for a spec to become an
OASIS standard, implementations are required as a prerequisite and thus we
need to be careful in our selection of a
technology/spec).
Insofar that it makes sense to
tackle Req 011-014 along with Req XXX (Semantic Search), let's do so, though
I personally think that we need to split the discussion so that we can get
as much parallel activity as possible.
What makes sense to me for the
immediate future is to get a common understanding of OWL (and other relevant
options) and evaluate it against our taxonomy requirements and areas such as
inheritance, equivalencies and relationship between values of given value
sets. As a next step (perhaps as a parallel activity) we should embark on
defining our Semantic Search requirements;
document them; prioritize them; and identify
solutions.
Proposed
Agenda
I recommend that the agenda for Monday's call consist of a
discussion of OWL and its
applicability to Req 011-014. First on the
agenda should be a presentation by Massimo and Katia (they've offered to do so) to introduce
us to OWL.
As a next step and later meeting, OWL-S can be
reviewed as part of discussing the Semantic Seach Requirement (Req-XXX); the
outcome of this work needs to be a requirements document that we evaluate
and prioritize separately.
Max: Do you agree with this
course of action?
Massimo/Katia: would
you be prepared to give a presentation on OWL on
Monday?
Regards,
Luc
Luc Clément
Secretary, OASIS UDDI Spec
TC
Cell:
425.941.0150
[1] UDDI V4 Requirements List, Section
2.7, RQ 011 - 014 - Taxonomy Items, http://www.oasis-open.org/apps/org/workgroup/uddi-spec/download.php/5050/UDDI%20v4%20Requirements%20List.htm#_Toc60926062
[2] Taxonomy items: API, modeling,
format, value sets, i18n, etc., http://www.oasis-open.org/apps/org/workgroup/uddi-spec/download.php/4065/uddi-spec-tc-rq011-14-20031030.doc