RE: NEW issue 6 - defining semantic relationships

One of the immediate applications that occurs to me for RDF thesauruses is
in mapping between discrete taxonomies. Whatever approach to defining
relationships is adopted, I hope it will be straightforward to take a tree
like the Open Directory website categorization [1] and be able to map across
to corresponding terms in another hierarchy like the Topic Exchange [2], and
carry out merge-like operations (I say merge-like because the structures may
be incompatible, and class/subclass relationships may only work in localized
regions of the trees).

Cheers,
Danny.

[1] http://dmoz.org/
[2] http://topicexchange.com/

> -----Original Message-----
> From: public-esw-thes-request@w3.org
> [mailto:public-esw-thes-request@w3.org]On Behalf Of Miles, AJ (Alistair)
>
> Sent: 11 November 2003 16:33
> To: 'public-esw-thes@w3.org'
> Subject: NEW issue 6 - defining semantic relationships
>
>
>
> I've added this issue to the RDF Thesaurus ESW Wiki.
>
> http://esw.w3.org/topic/RdfThesaurus?action=show
>
> Short summary:
>
> Issue 6 - Defining semantic relationships
>
> Description: A thesaurus consists of concepts, labels for concepts, and
> semantic relationships between concepts. A semantic relationship is a
> relationship of meaning. Most thesauri use a similar set of semantic
> relationships, which they label 'broader' 'narrower' and 'related'.
>
> Problem 1: 'broader/narrower' means different things in different
> thesauri.
> In some thesauri it means strictly class-subsumption. In other thesauri it
> can mean either is-a, instance-of, or part-of. Also 'related' is not
> consistently used. For example some thesauri model part-of relations with
> 'related', others use 'broader/narrower'
>
> => We must invent some mechanism for providing clear definitions
> of semantic
> relationships, and for removing any scope for ambiguity.
>
> Problem 2: some thesauri have semantic relations other than
> 'broader/narrower' and 'related'. Some overcome the 'broader/narrower'
> fuzziness by using 'BTI', 'BTG' and 'BTP', which stand for
> 'broader-term-instantive' 'broader-term-generic' and
> 'broader-term-partitive' respectively. In others there are custom
> relationships like 'related-broader'.
>
> => We must provide some mechanism by which users can extend the given
> relationship set and define their own semantic relations.
>
> .....
>
>
> Alistair Miles
>
> CCLRC - Rutherford Appleton Laboratory
> Building R1 Room 1.60
> Fermi Avenue
> Chilton
> Didcot
> Oxfordshire OX11 0QX
> United Kingdom
>
> Email:        a.j.miles@rl.ac.uk
> Telephone: +44 (0)1235 445440
>
>

Received on Tuesday, 11 November 2003 12:02:20 UTC