Re: [TaxoCoP] "Parametrics" related to taxonomy
- Ah, okay. So it is related or synonymous with faceted searching. Boy, keeping up with the terminology in this business is more challenging than creating the taxonomies. Thanks a lot.
Patrick Lambe <plambe@...> wrote:I have seen search vendors using"parametric search" as a synonym for faceted search - ie searching on a combination of attributes from different facets.POn 06 Apr 2007, at 7:42 AM, Ari Wackernah wrote:
Finding fabulous fares is fun.
Let Yahoo! FareChase search your favorite travel sites to find flight and hotel bargains.
- Wow, Lisa.That's a good one. Enough meat in those slides for a full-semester course.Among many other things, I like Obrst's distinction between "weak" and "strong" taxonomies.Thanks!Phil
From: TaxoCoP@yahoogroups.com [mailto:TaxoCoP@yahoogroups.com] On Behalf Of Lisa Colvin
Sent: Monday, April 16, 2007 1:06 PM
Subject: RE: [TaxoCoP] Nomeclature - ontologies
Great discussion! Ontolog, an ontology community of practice, issued a questionnaire to understand how people use and understand the word “ontology” which will be discussed at the upcoming Ontology Summit as part of NIST interoperability week (http://ontolog. cim3.net/ cgi-bin/wiki. pl?OntologySummi t2007). There have been many discussions in the group as to what constitutes an ontology.
In my experience, I’ve found it best to use the minimum appropriate level of knowledge representation language that is necessary for completeness and extensibility. The more expressive the language, the greater ability you have to define a more precise semantics. However, creating an ontology in a more expressive language is more difficult to maintain from a resource perspective and also may have challenges with respect to computability.
The best way I’ve found to express the ontology continuum (from taxonomies through logical theories) is along the axis of semantic expressivity as described by Leo Obrst (Mitre). You can find an overview presentation under the title “Ontology Spectrum, Semantic Models”(ontolog.cim3. net/.../presenta tion/LeoObrst_ 20060112/ OntologySpectrum SemanticModels- -LeoObrst_ 20060112. ppt) In particular, slide 9. This makes sense to me as an ontologist, but I’m curious as to whether this resonates with the Taxonomy community. (disclosure – I used to work for Leo.)
:::Lisa Dawn Colvin, Ontologist:: :
From: TaxoCoP@yahoogroups .com [mailto: TaxoCoP@yahoogroups .com ] On Behalf Of Phil Murray
Sent: Saturday, April 14, 2007 6:04 AM
To: TaxoCoP@yahoogroups .com
Subject: RE: [TaxoCoP] Nomeclature
Sowa is one of those people whose work you discover and then write a note to yourself that reads something like, "Find more stuff by this guy." Same for Guarino and McGuinness.
Thanks for the reminder. I'll definitely get his book.
You can find some of his more recent comments with a Google search that includes Sowa "semantic web" "old fashioned". As someone who has been involved in hypertext since 1986 and started learning about SGML in 1987, I am sympathetic with many of his views.
Although I'm a big fan of RDF, one of the questions I have posed explicitly to our development team is, "What limitations does RDF impose on our tools for gathering, representing, integrating, and managing knowledge?"
From: TaxoCoP@yahoogroups .com [mailto: TaxoCoP@ yahoogroups. com ] On Behalf Of Bob Doyle
Sent: Friday, April 13, 2007 3:28 PM
To: TaxoCoP@yahoogroups .com
Subject: Re: [TaxoCoP] Nomeclature
Very nice article with a good historical perspective. But remember that the Semantic Web and Web Ontology Language (OWL) still lie in the future for Smith and Welty.
Have you read John Sowa's book, Knowledge Representation? It was published about the same time as this article (2000).
Sowa provides a similar philosophical background and his work owes a lot to C.S. Peirce.
On 4/13/07, Phil Murray <phil.murray@ aelera.com> wrote:
The best thing I have read recently on ontologies -- and I think ontologies are highly relevant to this discussion because they represent part of the continuum of knowledge-represent ation systems -- is Barry Smith and Christopher Welty, " Ontology: Towards a New Synthesis" (2001) http://www.cs. vassar.edu/ faculty/welty/ papers/fois- intro.pdf .
Smith and Welty's delightfully brief and cheeky look at the origins of computer ontologies has the added benefit of pointing out that several different domains have been forced to turn to knowledge-organizat ion systems to solve information management problems. This is not "the revenge of the librarians" (Peter Morville) but the inevitable consequence of the proliferation of computers and free global self-publishing.
For me, hierarchy is absolutely fundamental to my understanding of taxonomy. I'm developing a product classification scheme for an e-commerce system. My core vocabulary consists of terms arranged in a parent-child hierarchy. There are "type of" (genus-species) relationships and "part of" (whole part) relationships. Sometimes a parent has two distinct sets of children--one set of "type" children, and another set of "part" children. This is the mission-critical part of my taxonomy.
Associative relations are secondary in importance. These are the "use for" and "related to" relationships among terms. It's my understanding that the addition of associative relationships is what turns a "taxonomy" into a "thesaurus". Based on some of the responses in this thread, I'm beginning to think my understanding is totally wrong. My original point was that I use the word "taxonomy" to describe the combined hierarchical and associative relationships, even though I know this means "thesaurus". But now I'm even more confused! :-)
Jame M. mentioned "ontologies" . I agree that this is a totally different beast, not to be confused with taxonomy. I use the term "ontology" to mean the rules (grammar) for using a taxonomy in a specific application. My taxonomy is system-agnostic. The functional implementations are ontologies. Thoughts?
This e-mail message and its attachments may contain confidential or legally privileged information and are for the sole use of the designated recipient(s) . If you are not a designated recipient of this message, please do not read, copy, use, or disclose this message or its attachments; and notify the sender by replying to this message, and delete or destroy all copies of this message and attachments in all media. Thank you.
Editor In Chief, CMS Review - http://www.cmsrevie w.com
Former Technology Advisor, CM Pros - http://www.cmprofes sionals.org/ membership/ cm-profiles/ bob-doyle
Contributing Editor, EContent Magazine -
http://www.econtent mag.com/About/ AboutAuthor. aspx?AuthorID= 155
President and CEO, skyBuilders - http://www.skybuild ers.com
77 Huron Avenue
Cambridge , MA 02138
Tel: +1 617-876-5676 Skype:bobdoyle