Re: Lee, Eugene, Grant, Jack?

From: Jack Park (jackpark@verticalnet.com)
Date: Fri Jan 19 2001 - 06:42:45 PST


I see reference to somebody named Jack and I perk up. Howard and I plan to
be at the meeting on monday. Additionally, we plan to be at a discussion at
SRI today (2:30) as pasted below:

Title: Ontology acquisition, integration, and merging

Speaker: Natasha Fridman Noy, Stanford University Medical Informatics
Group

Date: Friday Jan. 19th

Time: 2:30pm - 3:45pm NB Unusual day and time

Place EJ228 SRI International

________________

In recent years ontologies have become essential parts of many
applications. Today even the plain hypertext documents on the World-Wide
Web are turning into collections of small knowledge bases. The WWW
consortium is developing the Resource Description Framework (RDF)a
language for encoding knowledge on the Web pages that will be
understandable to electronic agents searching for information.

In our laboratory, we have developed Protege--a graphical and
easy-to-use ontology-editing and knowledge-acquisition environment. It
has an extensible architecture that allows other developers to plug in
components to perform various knowledge-acquisition and
knowledge-analysis tasks. Protege has a flexible OKBC-compatible
knowledge model and an extensible architecture. These two features allow
developers to adapt Protege-2000 to work as an ontology-editing
environment for other knowledge-representation systems and languages. We
have adapted it to become an editor for ontologies represented in RDF,
for example. In this process we reconciled the differences between the
knowledge models of Protege-2000 and RDF declaratively at the knowledge
level and made them transparent in the user interface.

There is also a wealth of reference information available electronically
that is not represented in formal ontologies and knowledge bases
however. These resources often include ontological information but are
not explicit ontologies themselves. The Unified Medical Language System
(UMLS) is one such resource. Even though the organization of knowledge
in UMLS may be different from that in the ontology that a domain expert
is creating, a lot of information could be imported directly from UMLS
and then reorganized into the categories that are appropriate for the
user's task. We have developed a remote-access client that allows
experts who are developing and populating their knowledge bases in
Protege-2000 to import elements from such knowledge sources and
integrate them in the evolving ontologies directly. The variety of
knowledge sources that have been developed by various groups also leads
to the problem of merging the information in these sources. We have
developed Prompt--a semi-automatic tool for ontology merging. PROMPT
performs some ontology-merging tasks automatically and guides the user
in performing other tasks for which his intervention is required. PROMPT
also determines possible inconsistencies in the state of the ontology,
which result from the user's actions, and suggests ways to remedy these
inconsistencies. The PROMPT algorithm is based on an extremely general
knowledge model and therefore can be applied across various platforms.
Our formative evaluation showed that a human expert followed 90% of the
suggestions that PROMPT generated and that 74% of the total
knowledge-base operations invoked by the user were suggested by PROMPT.

_______________

Natasha is a research scientist at Stanford Medical Informatics Group.
She is involved in the following projects:

* Protégé
* PROMPT (formerly SMART)
* Foundational Model of Anatomy

Her research interests are ontology design, ontology management,
knowledge acquisition, knowledge representation, evaluation of
knowledge-based systems

A list of publications can be found at
http://smi-web.stanford.edu/people/noy/publications.html

_________________

For more information on the AI Seminar Series see
http://www.ai.sri.com/ai-seminars/

============================================================================
This message is intended only for the use of the Addressee(s) and may
contain information that is PRIVILEGED and CONFIDENTIAL. If you are not
the intended recipient, dissemination of this communication is prohibited.
If you have received this communication in error, please erase all copies
of the message and its attachments and notify postmaster@verticalnet.com
immediately.
============================================================================



This archive was generated by hypermail 2.0.0 : Tue Aug 21 2001 - 17:58:00 PDT