W3C home > Mailing lists > Public > public-esw-thes@w3.org > November 2009

Re: [Dbpedia-discussion] Using DBpedia resources as skos:Concepts?

From: Leonard Will <L.Will@willpowerinfo.co.uk>
Date: Thu, 5 Nov 2009 23:22:47 +0000
Message-ID: <Va5l8hPH518KFAwb@mail.willpowerinfo.co.uk>
To: Simon Spero <ses@unc.edu>
Cc: Pat Hayes <phayes@ihmc.us>, Alexandre Passant <alexandre.passant@deri.org>, Richard Cyganiak <richard@cyganiak.de>, dbpedia-discussion@lists.sourceforge.net, SKOS <public-esw-thes@w3.org>
On Thu, 5 Nov 2009 at 17:05:16, Simon Spero <ses@unc.edu> wrote
>On Thu, Nov 5, 2009 at 3:25 PM, Leonard Will <L.Will@willpowerinfo.co.uk>wrote:
>> As you can certainly think about grains or sand, the fall of Carthage, or
>> Mrs Obama, these fall within our definition of "concept". Perhaps some other
>> word could be found to express this better and avoid confusion with a
>> narrower definition such as "abstract concept", but the word "concept" is
>> widely used in the thesaurus literature, in order to make a distinction
>> between the thing that is thought about and the words that may be used to
>> label it.
>> My view of this is from the approach of the library / thesaurus / knowledge
>> organisation community and the ISO thesaurus standard working party, and I
>> cannot say definitively that the SKOS interpretation is the same - there
>> have been some erudite discussions here about the difference between a thing
>> and our thoughts about the thing, but from a practical point of view of
>> applying indexing terms to resources these seem unnecessary.
>Obligatory unicorn: http://www.ibiblio.org/fred2.0/wordpress/?p=30
>The SKOS working group explicitly rejected the  interpretation that a
>skos:Concept is something that a Document is about, but declined to provide
>an explicit alternative.
>There are practical implications for indexing that follow from this
>decision.  For example, the SKOS broader relationship is not transitive;
>this is hard to understand with a document based domain of interpretation.
>Without transitive BT relationships, standard indexing behaviors like upward
>posting, or assigning the most specific headings to a document are no longer
>possible  (or rather, give different results).
>Once one starts thinking extensionally this whole discussion becomes much
>easier ("Word and Subject?").

I don't want to start again the whole discussion about transitivity, but 
let me just work through your example to see that I understand it:

>For example:
>Everything that is-about something is a document.
>Everything that something is-about is a concept.


>Every generic-concept is a concept.
>Every named-individual-concept is a concept.

OK, by definition presumably, as these are just qualifications narrowing 
the general concept "concept".

>Every concept that has-associated-class K is a generic-concept.
>Every concept that has-associated-individual I is a

Presumably by "associated class" you mean two or more individuals, not 
necessarily individually identified, each of which fits within the scope 
of the concept.

>If A has-broader-term-generic B then A has-broader-term B.
>If A has-broader-term-instantive B then A has-broader-term B.

I.e. generic and instantive relationships are specific cases of the more 
general "broader / narrower" relationship. (Strictly speaking the 
relationships are between concepts rather than between terms, but we'll 
accept that that is what is meant because of the historical/conventional 
use of BT/NT, BTG/NTG etc. for these relationships.)

>If A has-broader-term-generic B then A is a generic-concept.
>If A has-broader-term-generic B then B is a generic-concept.

I.e. a generic relationship can exist only between classes, not between 
a class and an individual

>If A has-broader-term-instantive B then A is a named-individual-concept.
>If A has-broader-term-instantive B then B is a generic-concept.

I.e. an instantive relationship can exist only between an individual and 
a class

>If A has-broader-term-generic B and
>   A has-associated-class X and
>   B has-associated-class Y
> then X subclassOfs Y.

OK, X is a subclass of Y

>If A has-broader-term-instantive B and
>   A has-associated-individual X and
>   B has-associated-class Y
> then X types Y.

I'm not sure whether you are using "types" with a formal meaning here, 
but I take it that you mean that X is a member of the class Y. (Sounds 
as though this should be Y types X, but I don't know the convention.)

>If a concept A has-broader-term B and
>   B has-broader-term C
> then A has-broader-term C.
>If a concept A has-broader-term-generic B and
>   B has-broader-term-generic C
>then A has-broader-term-generic C.
>If a concept A has-broader-term-instantive B and
>   B has-broader-term-generic C
> then A has-broader-term-instantive C.

In other words, you can have a branch of a tree made up of generic 
relationships, possibly with an instantive (or partitive) relationship 
at the leaf end, and these are all transitive, being an instantive (or 
partitive) relationship if the end concept is included and generic 

You can't have an instantive relationship other than at the end, because 
an individual by definition does not have members.

It's only when you have a partitive relationship that is not the end 
relationship of the branch that non-transitivity occurs.

Since in standard thesaurus structures partitive relationships are 
restricted to a few special cases such as places and organisational 
structures, this is not usually a problem in practice.

Have I interpreted you correctly?

Leonard Will
Willpower Information     (Partners: Dr Leonard D Will, Sheena E Will)
Information Management Consultants            Tel: +44 (0)20 8372 0092
27 Calshot Way                              L.Will@Willpowerinfo.co.uk
ENFIELD                                Sheena.Will@Willpowerinfo.co.uk
EN2 7BQ, UK                            http://www.willpowerinfo.co.uk/
Received on Thursday, 5 November 2009 23:23:37 UTC

This archive was generated by hypermail 2.4.0 : Friday, 17 January 2020 19:45:59 UTC