Re: [xtm-wg] Common Assumptions

[Murray Altheim <altheim@eng.sun.com>]

univers immedia wrote:
> 
> I'm glad this type of question is raised.
> 
> The very question is : what in a TM carries ontology, and what are the
> processes able to verify the alignment of ontologies before merging ? What
> can be system-driven and what needs some sort of human agreement ? This is
> a complex and largely open issue, and with certainly no one-for-all
> solutions.
> 
> Agreement on identity of topic subjects leads us back to the debate about
> baseName-based vs subjectIndicator-based merging, and the lack of a general
> theory of scope and context. [...]
>
> Agreement on identity of [meaning of] associations and roles is a more
> tricky but crucial issue, not really addressed so far, and maybe linked
> with the other debate on reification, since the identification/merging
> process of two associations maybe needs a proper identification/merging
> process of the topics that reify them ...
> 
> Anyway all that confirms that TM technology development will be of limited
> use without coordinated development of public ontologies strongly grounded
> in (stable) PSI.

Actually, I wasn't advocating the requirement of public ontologies, but
that we address (in the specification) the issue of semantic agreement
between *all* components in topic maps prior to merging, not just topics
based on subject. As you point out, name-based merging requires that 
there be agreement on both name and scope, as well as what the scope
provides semantically. While in the abstract this seems somewhat 
straightforward, I think it will be profoundly difficult to implement
programmatically. 

For example, to return to ontologies and the like, it would be exceedingly
valuable to map the relations between a common ontology like Cyc and the 
US Library of Congress subject headings, but I see no way to perform this
via computer (at this time). The manual (ie., human) process of mapping 
over 320,000 subjects to terms in a similarly-sized ontology would be 
daunting at best, and would even then only serve to provide the map based
on the context in which the human judged equivalence. This is a vexing
problem that is only *highlighted* in this example, but shows up in every
merge process.
 
Murray

...........................................................................
Murray Altheim, SGML/XML Grease Monkey     <mailto:altheim&#64;eng.sun.com>
XML Technology Center
Sun Microsystems, 1601 Willow Rd., MS UMPK17-102, Menlo Park, CA 94025

      In the evening
      The rice leaves in the garden
      Rustle in the autumn wind
      That blows through my reed hut.  -- Minamoto no Tsunenobu

------------------------ Yahoo! Groups Sponsor ---------------------~-~>
eGroups is now Yahoo! Groups
Click here for more details
http://click.egroups.com/1/11231/0/_/337252/_/980793564/
---------------------------------------------------------------------_->

To Post a message, send it to:   xtm-wg@eGroups.com

To Unsubscribe, send a blank message to: xtm-wg-unsubscribe@eGroups.com