Add class KnowledgeConcept

[rjyounes]

gist:KnowledgeConcept
	a owl:Class ;
	rdfs:subClassOf gist:IntellectualProperty ;
	skos:definition "These are abstract concepts that arise from the distillation of experience.  Similar to Categories but different in that categories are typically simple tags, and these items contain rich structures."^^xsd:string ;
	skos:example """Gene, Protein, Chemical,Disease, Customer Intent (in the customer service sense of the word), Commodity Grade, Subject Matter (law), Method, Incident, Problem, ProssibleCause, Topic, Role, Industry (IKG).

Most domains will have a few of these classes as sub classes of Knowledge Concept."""^^xsd:string ;
	skos:prefLabel "Knowledge Concept"^^xsd:string ;
	skos:scopeNote """Some knowledge is about specific instances that already exist in the knowledge graph.  We may have knowledge that Judge Jones is more lientient on repeat offenders in the morning, we may know that the earliest killing frost in Fort Collins is the last week of September.

These fit the broader definition of knowledge, the distillation of experience.  But these do not require any new instances.  Judge Jones and Fort Collins already exist and have a place in our knowledge graph.

But some knowledge requires that we synthesize new instances in order to have a place to hang knowledge.  A disease isn't a tangible thing, like a person or a building.  A disease is a prediction that a persons health will decline in a predictable way (without treatment and with treatment).  The number of diseases continues to grow,  as we collectively learn more and more fine grain distinctions. Lung Cancer used to be a disease but now we have dozens of fine grained distinctions, such as non lymphoma small cell fusiform is different from alveolar adenocarcinoma because we now know the prognosis and treatment are different.

This superclass is meant as a place for subclasses (such as :Disease) which will have the instances that represent the foci of the knowledge we have acquired."""^^xsd:string ;
	.

[rjyounes]

See comments on issue #946.

[uscholdm]

@rjyounes @mkumba
This is very interesting.

[knowledge concepts are] similar to Categories but different in that categories are typically simple tags, and these items contain rich structures.

If this is the case, then perhaps the similarity can be expressed semantically. In English a disease is a categorization of a set of conditions about one's health. Associated with a disease are predictions that a persons health will decline in particular ways (without treatment and with treatment).

Gene, Protein, Chemical,Disease, Customer Intent (in the customer service sense of the word), Commodity Grade, Subject Matter (law), Method, Incident, Problem, ProssibleCause, Topic, Role, Industry (IKG)

Pretty much all of these can be viewed as categorizing something - indeed, several of these showed up in the ontologies I built and were modeled as subclasses of gist:Category

Maybe Category is a special less interesting kind of KnowledgeConcept, one w/o structure often used as a tag.

[mkumba]

It's kind of what I'm thinking, these knowledge concepts are very like categories, but with a lot of complex structure to them. I kind of like that categories are mostly terminal, immutable tags with little new to say about them, but maybe knowledge is just category promoted Dave McComb, CEO https://www.semanticarts.com On Mar 4, 2025 at 12:31 PM -0700, Michael Uschold ***@***.***>, wrote: @rjyounes<https://github.com/rjyounes> @mkumba<https://github.com/mkumba> This is very interesting. [knowledge concepts are] similar to Categories but different in that categories are typically simple tags, and these items contain rich structures. If this is the case, then perhaps the similarity can be expressed semantically. In English a disease is a categorization of a set of conditions about one's health. Associated with a disease are predictions that a persons health will decline in particular ways (without treatment and with treatment). Gene, Protein, Chemical,Disease, Customer Intent (in the customer service sense of the word), Commodity Grade, Subject Matter (law), Method, Incident, Problem, ProssibleCause, Topic, Role, Industry (IKG) Pretty much all of these can be viewed as categorizing something - indeed, several of these showed up in the ontologies I built and were modeled as subclasses of gist:Category Maybe Category is a special less interesting kind of KnowledgeConcept, one w/o structure often used as a tag. — Reply to this email directly, view it on GitHub<#1198 (comment)>, or unsubscribe<https://github.com/notifications/unsubscribe-auth/AAAGJPWGARPHFSIKWVPXDS32SX5QHAVCNFSM6AAAAABXAXZ4MGVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDMOJYGY3DGNRYHE>. You are receiving this because you were mentioned.Message ID: ***@***.***> [uscholdm]uscholdm left a comment (semanticarts/gist#1198)<#1198 (comment)> @rjyounes<https://github.com/rjyounes> @mkumba<https://github.com/mkumba> This is very interesting. [knowledge concepts are] similar to Categories but different in that categories are typically simple tags, and these items contain rich structures. If this is the case, then perhaps the similarity can be expressed semantically. In English a disease is a categorization of a set of conditions about one's health. Associated with a disease are predictions that a persons health will decline in particular ways (without treatment and with treatment). Gene, Protein, Chemical,Disease, Customer Intent (in the customer service sense of the word), Commodity Grade, Subject Matter (law), Method, Incident, Problem, ProssibleCause, Topic, Role, Industry (IKG) Pretty much all of these can be viewed as categorizing something - indeed, several of these showed up in the ontologies I built and were modeled as subclasses of gist:Category Maybe Category is a special less interesting kind of KnowledgeConcept, one w/o structure often used as a tag. — Reply to this email directly, view it on GitHub<#1198 (comment)>, or unsubscribe<https://github.com/notifications/unsubscribe-auth/AAAGJPWGARPHFSIKWVPXDS32SX5QHAVCNFSM6AAAAABXAXZ4MGVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDMOJYGY3DGNRYHE>. You are receiving this because you were mentioned.Message ID: ***@***.***>

[rjyounes]

gist dev team meeting 5/8:

Rebecca: There's a relationship to punning. We want to say something about the concept of dog.
Michael: A particular gold bar is an instance of physical identifiable item. If you want to say something about gold itself, then gold is a knowledge concept.
Rebecca: You have to be able to relate the particular instance of gold to what we know about the concept of gold. How would we do that?

Rebecca:
Class Dog
Instances: Rover, Fido
KnowledgeConcept Dog - typically has 4 legs, is a mammal, ...
How do these relate to one another? Is it helpful to introduce the knowledge concept?

Straw poll:
Leaning yes: 1
Leaning no: 2
Undecided: 4

Michael: In English these are categories.
Rebecca: I don't think of these examples as categories.

DECISION: Continue discussion when Dave is present.

[JonathonGist]

I like the idea of adding in gist:KnowledgeConcept. I can see some use for it in my personal work, outside of my current employer.

[mdebellis]

I like the idea of adding in gist:KnowledgeConcept. I can see some use for it in my personal work, outside of my current employer

There is already a class in SKOS called Concept. Is there a good reason to have something called a KnowledgeConcept that is different from skos:Concept?

…

On Thu, May 8, 2025 at 1:20 PM JonathonGist ***@***.***> wrote: *JonathonGist* left a comment (semanticarts/gist#1198) <#1198 (comment)> I like the idea of adding in gist:KnowledgeConcept. I can see some use for it in my personal work, outside of my current employer. — Reply to this email directly, view it on GitHub <#1198 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AD7TPBFDFGLJ33A4IN7MHD325O4BBAVCNFSM6AAAAABXAXZ4MGVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDQNRUGE4TANBQHA> . You are receiving this because you are subscribed to this thread.Message ID: ***@***.***>

[JonathonGist]

There is already a class in SKOS called Concept. Is there a good reason to
have something called a KnowledgeConcept that is different from
skos:Concept?

It seems to me that skos:Concept is the more generic, non-differentiated superclass that can encompass both gist:Category and gist:KnowledgeConcept. These 2 gist classes provide more nuanced semantics that have merit on their own, in my eyes. But I'm not part of the gist team nor a highly-experienced ontologist.

As I archive my artistic works described in #1189, I would like to create gist:KnowledgeConcept individuals and link each work to one or multiple of these. This is a very exciting use case for my personal work. If you don't bring it fully into gist, I may borrow it for my own ontology (an extension of gist).

[mdebellis]

It seems to me that skos:Concept is the more generic, non-differentiated superclass that can encompass both gist:Category and gist:KnowledgeConcept. These 2 gist classes provide more nuanced semantics that have merit on their own, in my eyes. But I'm not part of the gist team nor a highly-experienced ontologist.

That doesn't sound like an actual use case. Where would you need the difference between a Concept and a KnowledgeConcept? If you have the concept defined in an ontology that is one of the most widely used vocabularies then unless there is a good argument not to, I think it makes more sense to just use that class. One of my favorite parts of Uschold's OWL for the Enterprise is: It is easy to be lured into thinking that more is better. It’s not. Smaller

and simpler ontologies are easier to understand and use, and more computationally efficient as well. A good principle to use is “when in doubt, leave it out.”

Uschold, Michael. Demystifying OWL for the Enterprise (p. 203). Morgan & Claypool Publishers. Kindle Edition. I think that's true for ontologies in general and even more so for upper models. Michael https://www.michaeldebellis.com/blog

…

On Thu, May 8, 2025 at 4:19 PM JonathonGist ***@***.***> wrote: *JonathonGist* left a comment (semanticarts/gist#1198) <#1198 (comment)> There is already a class in SKOS called Concept. Is there a good reason to have something called a KnowledgeConcept that is different from skos:Concept? It seems to me that skos:Concept is the more generic, non-differentiated superclass that can encompass both gist:Category and gist:KnowledgeConcept. These 2 gist classes provide more nuanced semantics that have merit on their own, in my eyes. But I'm not part of the gist team nor a highly-experienced ontologist. — Reply to this email directly, view it on GitHub <#1198 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AD7TPBF3JPZZELQ4CSU3Y5325PQ7DAVCNFSM6AAAAABXAXZ4MGVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDQNRUGY3DQNJQHE> . You are receiving this because you commented.Message ID: ***@***.***>

[JonathonGist]

For an upper ontology, I can see that gist:KnowledgeConcept may be too specific. For my need, I want to distinguish between gist:Category and skos:Concept with :KnowledgeConcept as written at the top of this issue.

For my personal ontology I plan to use it as-is, extending gist (if gist:KnowledgeConcept isn't added to a future release). I have a few individuals of :KnowledgeConcept to add to my knowledge graph. They would be distillations of my experience gained over the years; and specific statements or linkages to presentations I have done on domain knowledge. (How to record nature sounds, how to be in nature and why to be there; how to listen better.) I don't have the time yet to flesh out these use cases.

I look forward to seeing what the gist team does with gist:KnowledgeConcept! Thank you for allowing me to share.