- From: Leonard Will <l.will@willpowerinfo.co.uk>
- Date: Sat, 7 Jan 2023 17:29:01 +0000
- To: Montens Pieterjan <Pieterjan.Montens@curia.europa.eu>
- Cc: "public-esw-thes@w3.org" <public-esw-thes@w3.org>
- Message-ID: <cf12cf8f-221f-134d-1373-80bf652e74a7@willpowerinfo.co.uk>
The definition of a concept often cannot be assumed from its label or labels or its relationships; it needs a scope note to say how the concept is defined and how it is to be understood within the context of the knowledge organisation scheme in which it occurs. If MorningStar and EveningStar are different concepts, their scope notes should spell out the criteria by which they can be distinguished. They may be sibling concepts, with a parent (broader) concept such as "stars" (or "planets" or "astronomical phenomena" or whatever). There is then a semantic relationship between them because of this hierarchical structure. Leonard Will On 06/01/2023 08:58, Montens Pieterjan wrote: > > I’d like to elaborate on Tom’s answer, who is absolutely right : in a > vocabulary you want to distinguish the signified (the concept behind > the labels) and the signifiers (the labels that represent the abstract > concept). The root identifier of a concept is preferably an abstract > key without meaning (numbers, letters, UUID or a persistent ID schema) > to which you attach labels (SKOS-XL extends the description and > linking of lexical entities). Under this definition, two identical > concepts in the same concept scheme would introduce confusion and > inconsistency. > > All of this because labels change in space and time : if EveningStar > is nothing but another name for MorningStar and represents the same > thing or concept, using labels is the way to go. If at one point > EveningStar becomes a closely related but different entity from > MorningStar, they represent different concepts and could be related > (like in skos:related : there’s an associative semantic relation, but > not in a hierarchical sense). And if in the end the concept behind > EveningStar becomes entirely separated from the one behind > MorningStar, there simply are no more semantic relations between them > (might be useful to throw in a history note though). > > The Customer-Client question is to be solved by your own needs and > practical limits : is there a business case where you need to > distinguish those roles, or are they, as far as you are concerned, > used interchangeably ? > > Pieterjan Montens > > *From:*Tom Morris <tfmorris@gmail.com> > *Sent:* vendredi 6 janvier 2023 03:57 > *To:* Michael DeBellis <mdebellissf@gmail.com> > *Cc:* public-esw-thes@w3.org > *Subject:* Re: skos:closeMatch and skos:exactMatch > > On Thu, Jan 5, 2023 at 6:14 PM Michael DeBellis > <mdebellissf@gmail.com> wrote: > > If I have two synonym Concepts in two different ConceptSchemes: > ex1:TheMorningStar and ex2:TheEveneningStar I can say > ex1:TheMorningStar skos:exactMatch ex2:TheEveningStar. But what if > they are in the /same /Concept Scheme? If I want to say > ex1:TheMorningStar is identical to ex1:TheEveningStar there > doesn't seem to be a way to do that. Similarly if I want to say > they are close but not quite exact synonyms (e.g., ex1:Customer > and ex1:Client) What am I missing? > > You wouldn't have two identical concepts in the same scheme. Synonyms > can be done using altLabels on the same concept. > > Tom > -- Leonard Will Thirlmere House, 29 Stanger Street, KESWICK, CA12 5JX Mobile telephone: 07905918640
Received on Saturday, 7 January 2023 17:29:20 UTC