[Top] [All Lists]

Re: [ontolog-forum] Taxonomies and Ontologies

To: "[ontolog-forum] " <ontolog-forum@xxxxxxxxxxxxxxxx>
From: Chris Menzel <cmenzel@xxxxxxxx>
Date: Wed, 13 Jul 2005 16:17:03 -0500
Message-id: <20050713211703.GE903@xxxxxxxx>
On Wed, Jul 13, 2005 at 08:55:28AM -0700, Jayne E Dutra wrote:
> I am looking for material regarding the relationship between
> ontologies and taxonomies.     (01)

Well, I suppose my own take counts as "material" of a certain possibly
useless and annoying sort. :-)  But to make at least a gesture in
someone else's direction, while I find a lot of his stuff quirky, I
think John Sowa's paper "Building, Sharing, and Merging Ontologies"
(http://www.jfsowa.com/ontology/ontoshar.htm) is quite helpful and
informative.  As for me:    (02)

Short answer:  A taxonomy is a type of ontology that describes a
hierarchy of classes.    (03)

Long, soporific, needlessly technical logician's answer:  In my view, an
ontology is, most fundamentally, a set of statements in a syntactically
well-defined, knowledge representation language.  Typically, for a given
ontology, the KR language in which the ontology is expressed is
"interpreted", in the sense that the language has an associated
"intended" domain (e.g., California wines, classes of automobiles, etc)
and its basic constructs all have associated "intended" meanings (e.g.,
the predicate "RedWine" indicates the property of being a red wine).
Any well-defined KR language should be translatable into (perhaps a
sublanguage of) a first-order language, so, theoretically, an ontology
is an axiomatic theory in an (interpreted) first-order language.    (04)

A taxonomy, then, is a special case of an ontology describing a
hierarchy of classes (or "categories") in some domain of interest.  A
"pure" taxonomy provides nothing more than a description of the
hierarchy; that is, a pure taxonomy supplies only a vocabulary of class
names and a collection of parent-child, or "inheritance", relationships
holding among the named classes.      (05)

Graphical languages provide the most vivid representations of pure
taxonomies, of course, but are easily translated into more traditional
first-order equivalent (notably, where the arrows, say, of the graphical
language expressing the "parent-child" relation between classes in the
hierarchy are replaced by a binary predicate symbol) .  Once you start
adding further axioms imposing more detailed, nonstructural constraints
on the classes in a hierarchy (and their instances) you move away from a
"pure" taxonomy in the direction of a general ontology.  Thus, an
ontology might -- but, in general, need not -- be based upon a pure
taxonomy.    (06)

Or so it all seems to me. :-)    (07)

Chris Menzel    (08)

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    (09)

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