[Top] [All Lists]

Re: [ontolog-forum] April 20 session on tagging ontolog content

To: nicolas.rouquette@xxxxxxxxxxxx, "[ontolog-forum] " <ontolog-forum@xxxxxxxxxxxxxxxx>
Cc: ontolog-forum@xxxxxxxxxxxxxxxx
Cc: ontolog-forum-bounces@xxxxxxxxxxxxxxxx
From: dbedford@xxxxxxxxxxxxx
Date: Thu, 6 Apr 2006 16:34:32 -0400
Message-id: <OFD6572349.9C2D135B-ON85257148.006ED751-85257148.0071066C@xxxxxxxxxxxxx>
Nicolas,    (01)

Yes, I believe that ultimately there are architectures and processes which
support 'ontologizing' content.  I think we are probably all working with those
now.   I'd like us to actually think about the idea of 'tagging' in a slightly
more logical way.  I have actually worked with many tools over the past 20 years
that variously tag, index, categorize, classify, summarize, etc.   The challenge
is to fit the type of technology to the nature of the entity and the
application.   'Tagging' will always means something different in a new context.    (02)

Next week I'll walk through quickly and simply the logical steps we've taken to
create an ontology - independent of tools, architectures and processes.   If I
can succeed at showing everyone how all of their views fit into a common
framework that will set us up to move forward.   My premise is that an actual
ontology may be complex, but a good ontology should always be grounded in a
simple reference model.    (03)

We actually began our work at an entity models level - looking at key entities,
and then identifying the attributes of those entities which supported all kinds
of use.   This is also a dynamic process - new uses are always emerging and the
requirements need to be integratable into the ontology model.  There is no one
final set of attributes for an entity that serves for all time, there is no one
set of entities that serves for all time.  Beyond this, the semantics need to be
addressed at the attribute level or at the entity element level.  Defining the
semantics for attributes (here I include syntax, behavior and semantics) are the
next step.  In some cases, the semantics are loose - as in Max's examples of
social characterization.  In otherr cases, the semantics need to be more robust.
The approach you take depends on how the attribute supports use and access.
And, the two approaches are not mutually exclusive.  After you've addressed
entities -- defining relationships among entities - similar to the axioms that
Cyc has defined -- is a next step.   Relationships also can include or define
context.    (04)

The challenge I have found working in the technical environment is that we tend
to jump too quickly into architectures and specifications without understanding
the purpose or nature of the application.   Projects which take the time to
understand the application from the user's perspective generally succeed and are
sustainable because they are based on a good basic reference model.    (05)

Best regards,
Denise    (06)

Message Archives: http://ontolog.cim3.net/forum/ontolog-forum/
Shared Files: http://ontolog.cim3.net/file/
Community Wiki: http://ontolog.cim3.net/wiki/ 
To Post: mailto:ontolog-forum@xxxxxxxxxxxxxxxx    (07)

<Prev in Thread] Current Thread [Next in Thread>