Content of the minimal vocabularies for restricting the CIDOC CRM Types

In the 53rd CIDOC CRM & 46th FRBRoo SIG meeting, the SIG returned to Issue 556 to establish what the next steps are and assign HW, since 496 has been resolved. 

So far: 

• TV has proposed a list of classes of the CRM, for which recommending a type would be useful.
• MD has suggested providing recommendations for (a) the types mentioned in properties the range of which is an instance of E55 Type (or a specialization thererof), and by extension all the .1 properties, (b) the types referred to (or implied) in the scope notes of classes and properties. 

Proposal: 

1. closely-read the scope notes of CRM classes/properties in order to determine what ontological distinctions can be implemented through the use of types,
2. go through the deprecated classes to determine whether their content and labels match classifications implemented in other known vocabularies (to check for the next meeting),
3. identify domain-specific vocabularies and standardized vocabularies that are relevant for the CRM.

Everyone in agreement, HW assigned to: 

• TV: check the classes he has identified for type-recommendations
• WS: check types of E4 Period (also settlements qua E7 Activity) and with vocabs from the DAI.
• MD: check the deprecated classes
• RN: skos vocabularies that they work with at the NFDI 

To be discussed again in the next Sig meeting. 

May 2022

Post by Martin Doerr (10 July 2022)

Dear All,

Attached a proposal how to develop a recommended vocabulary associated with the CRM, so far in order to refer to the meaning of deprecated classes.

My idea is to find good matches from AAT, if possible, or IFLA vocabularies.

Best,

Martin

Post by Melanie Roche (11 July 2022)

Dear Martin and all,

As far as E82 Actor Appellation is concerned, I would recommend against referring to FRAD as the model was superseded in 2017 by the IFLA LRM (which I don't think provides any good match either as such).
I am therefore not so sure it would be a good practice.

Best,

Mélanie.
 

Post by Thanasis Velios (11 July 2022)

Following this, I am also making a few recommendations on possible vocabularies based on my previous HW:

* E4: type of period → do not make recommendation

* E10: type of transfer of custody
    * legal responsibility → possible AAT term [ownership](http://vocab.getty.edu/page/aat/300055603)
    * physical possession → possible AAT term [possession (property right)](http://vocab.getty.edu/page/aat/300411616)

* E15: type of identifier assignment
    * "preferred identifier assignment" → CRM thesaurus

* E34: type of alphabet
    * List of script names → [ISO15924](https://www.unicode.org/iso15924/iso15924-codes.html)

* E57: type of material → do not make recommendation

* E58: type of unit
    * List of units → [ISO80000](https://www.iso.org/standard/30669.html)

* P3.1: type of encoding, type of note → do not make recommendation

* P14.1: type of role → do not make recommendation

* P16.1: type of mode of use → do not make recommendation

* P136.1: type of taxonomic role → do not make recommendation

* P19.1: type of use → do not make recommendation

* P62.1: mode of depiction → unclear what this is, no example

* P67.1: type of reference → do not make recommendation

* P138.1: mode of representation → do not make recommendation

* P69.1: type of association → do not make recommendation

* P102.1: type of title
    * child terms of AAT [titles](http://vocab.getty.edu/page/aat/300417193)

* P107.1: type of member → do not make recommendation

* P130.1: type of similarity → do not make recommendation

* P137.1: type of taxonomic role → do not make recommendation (well established in biology but not other disciplines)

* P139.1: type of alternative form → do not make recommendation

* P144.1: type of membership → do not make recommendation

* P189.1: type of approximation → unclear what this is, no example

In relation to deprecated classes, the only one I disagree with Martin is:

E40 Legal Body. AAT corporations (http://vocab.getty.edu/page/aat/300025969) is too focused on business.

All the best,

Thanasis

Post by Martin Doerr (11 July 2022)

Dear Thanassi,

I agree with all, except for:

On 7/11/2022 3:50 PM, Athanasios Velios via Crm-sig wrote:

Following this, I am also making a few recommendations on possible vocabularies based on my previous HW:

* E4: type of period → do not make recommendation

We need actually the interpretations of geopolitical units etc. I am working on this.

* E10: type of transfer of custody
    * legal responsibility → possible AAT term [ownership](http://vocab.getty.edu/page/aat/300055603)
    * physical possession → possible AAT term [possession (property right)](http://vocab.getty.edu/page/aat/300411616)

Sounds good, we need also an illegal possession concept..

* E15: type of identifier assignment
    * "preferred identifier assignment" → CRM thesaurus

* E34: type of alphabet
    * List of script names → [ISO15924](https://www.unicode.org/iso15924/iso15924-codes.html)

* E57: type of material → do not make recommendation

* E58: type of unit
    * List of units → [ISO80000](https://www.iso.org/standard/30669.html)

* P3.1: type of encoding, type of note → do not make recommendation

* P14.1: type of role → do not make recommendation

* P16.1: type of mode of use → do not make recommendation

* P136.1: type of taxonomic role → do not make recommendation

Actually I think we should recommend for biodiversity the GBIF terms, such as "holotype, lectotype" etc. someone to find the standard!

* P19.1: type of use → do not make recommendation

* P62.1: mode of depiction → unclear what this is, no example

"front", "back", "outline".....missing example is an ISSUE!

* P67.1: type of reference → do not make recommendation

* P138.1: mode of representation → do not make recommendation

"front", "back", "outline", or more.... missing example is an ISSUE!

* P69.1: type of association → do not make recommendation

* P102.1: type of title
    * child terms of AAT [titles](http://vocab.getty.edu/page/aat/300417193)

* P107.1: type of member → do not make recommendation

* P130.1: type of similarity → do not make recommendation

* P137.1: type of taxonomic role → do not make recommendation (well established in biology but not other disciplines)

Actually I think we should recommend for biodiversity the GBIF terms, such as "holotype, lectotype" etc. someone to find the standard!

* P139.1: type of alternative form → do not make recommendation

Question to IFLA members. Is there a good vocabulary?

* P144.1: type of membership → do not make recommendation

* P189.1: type of approximation → unclear what this is, no example

For instance, "centroid", "outer bound", "point within", "covering square"...missing example is an ISSUE!

In relation to deprecated classes, the only one I disagree with Martin is:

E40 Legal Body. AAT corporations (http://vocab.getty.edu/page/aat/300025969) is too focused on business.

Well, I have thought about this. The note says "commonly formed as business enterprises", not "always". I think the practical purpose of the construct was indeed "considered in law as legal persons having an existence and rights and duties distinct from those of the individuals who form them". This should include of course museums etc. that can do contracts, buy and sell. Since we have abandoned the concept, I'd regard the margin not covered by the AAT term as not important. Clearly, whatever we recommend for the deprecated classes, either it is our own term or a slight deviation from it. I prefer the latter, if we can adopt a well-established vocabulary.

Would you create a Google doc for this?

Best, 

Martin

Post by Pat Riva (12 July 2022)

Hi Martin, Thanasis,

So far the only IFLA vocabularies that are set up as linked data are here:

ISBD:

https://www.iflastandards.info/isbd

The ISBD vocabularies include several around types of content form or type of medium, but these are more restricted than any of these CRM classes.

Unimarc:

https://www.iflastandards.info/unimarc

Here there are some very specific vocabularies about certain material types.

There is really no vocabulary for the types of appellation relations for P139.1, in the IFLA space or others that I can think of. A bit surprising really.  I think that is because in library files there may be a couple of cases that are specifically coded and all other cases are lumped together as general derivation rather than being explained.

Similarly, I was sure we'd have a library vocabulary for P102.1 type of title, but actually it isn't a vocabulary but partially expressed in the element set. 

For the deprecated class P82 Actor Appellation, I agree with Melanie that we should not refer to FRAD which is superseded. But as for the other deprecated something-appellation classes, what the appellation names is expressed by the type of whatever the E41 is used for.

Pat

Post by Martin Doerr (12 July 2022)

Hi Pat,

Thank you for this analysis! So, we leave these cases without recommendations.

The "Actor Appellation" would be relevant for representing studies about etymology of person names without reference to a particular person, but we have not seen requests in CRM-SIG for encoding such studies so far. It may appear in genealogies or as auxiliary material for identifying historical persons, trying out typical alternative names.

The AAT distinguishes personal names:

ID: 300266386 Page Link: http://vocab.getty.edu/page/aat/300266386   Record Type: concept
	personal names (names, <names and related concepts>, ... Associated Concepts (hierarchy name))
Note: The name by which an individual person is identified or known, as distingushed from names for corporate bodies or other entities.

Does anyone have access to the latest AAT version in a way you can browse the hierarchy downwards? Curiously, the on-line page of the Getty vocabularies seems to provide no access to narrower terms. Is there an AAT concept "corporate names" or so?

In any case, "personal names" may be a concept to  recommend, once "corporate names" will not have an etymology.

Opinions?

Best,

Martin

In the 55th joint meeting of the CIDOC CRM and SO/TC46/SC4/WG9; 48th FRBR/LRMoo SIG meeting, the SIG began to revise the terms/vocabularies proposed to restrict the types pf deprecated classes.

Summary of decisions: 

• The functional role of the minimal vocabulary (text by MD) will undergo proof-reading (HW SdS) and will appear in the introductory section of the CIDOC CRM, right below "About Types", once SdS is done editing.
• (deprecated) E40 Legal Body is not equivalent to E47 Group.P2 has type: E55 Type {AAT corporations} OR {AAT: corporate bodies}. Either reconsider 'corporate bodies' or produce a scope-note that reflects the intended meaning.
• (deprecated) E45 Address can be replaced by E41 Appellation. P2 has type: E55 {AAT: street address}

The HW as a whole can be accessed here

HW: SdS to proofread MD's text.

Belval, December 2022

In the 56th joint meeting of the CIDOC CRM and ISO/TC46/SC4/WG9 &49th FRBR/LRMoo SIG, TV brought the SIG up to date with the developments in issue 556. At the time of the meeting things stood like that. 

• HW: SdS to proof-read it the text that MD has drafted regarding the functional role of the minimal vocabulary (pending decision from the 55^th SIG Meeting).
  The text is to appear in the introductory section of the CIDOC CRM, right after “About types” after SdS is done editing it.
• HW: MD to share the outcomes of the work he’s undertaken wrt the classification of geopolitical units with the SIG.

Summary of Decisions:

• The type recommendations for deprecated classes and the type recommendations for existing classes & typed properties, as well as the functional role of the minimal vocabulary will be accessible as a separate document.
  
  • The SIG is to reconsider whether (and how) it can appear as an appendix to the specification document.
  • The details of classes & typed properties to be rendered through types, plus the relevant types can be found in the below –see type recommendations for deprecated classes, existing classes & typed properties in the attached document
• Discuss the scope notes of typed properties in a new issue. P62.1, P67.1, P138.1, P189.1 are the most pressing (cryptic scope notes, lack of examples in some).

Crete, May 2023

Post by Martin Doerr (5 September 2023) 

Dear All,

Some years ago we analyzed place types from different gazetteers, with the focus on  such phenomena with a relevant spatiotemporal evolution:

I have made the following distinctions by abstracting from the Alexandria Gazetteer place types, according to the kind of phenomena that are responsible for their definition and identity and for avoiding possible polysemy of the same term/name. Similar place types appear in the TGN. Place types in Geonames should also be considered. An early version of place types from the INSPIRE standard appeared not to be as good.

A)      Distinct spaces defined by geomorphological forms (continents, islands, mountain ranges, water bodies, vulcanos)

B)      Distinct habitats defined by life form (kinds of vegetation etc.)

C)      Coherent, evolving human-maintained spaces (settlements, roads, areas formed by agriculture or other exploiation)

D)      Spaces defined by inhabitation/stay of a specific cultural group of people (town population, tribe, language group)

E)       Areas determined by execution of political power (Nation, country, administrative unit, protection zone)

F)       Possibly evolving areas defined by theoretical declaration motivated by scientific, social or political interests.

Wheras A), F) may characterize just spacetime volumes, B) through E) may characterize E4 Periods in the narrower sense.

It seems that only very few high-level abstractions are necessary to make a term like "Greece" or "Rome" unambiguous. Therefore the above may lead to a minimal vocabulary recommended by the CRM for E4 Period and Spacetime Volume
I attach the Alexandria terms.

Best,

Martin

Post by Dominic Oldman (5 September 2023)

Hi Martin,

Thanks for this.
 
When looking at this I started thinking about some projects I have been working on where, in one case, researchers are interested in the evolving landscape as markers for changes in the way people lived (and their conditions) and in another where administrative decisions (colonial) are based on hydrology surveys.

I think this is a useful practical abstraction.

Cheers,

Dominic

In the 57th CIDOC CRM & 50th FRBR/LRMoo SIG Meeting, the SIG reviewed (i) HW by MD (type restrictions for CRM STVs and Periods) and (ii) AK (mapping types specified for P136.1 & P62.1 to GBIF and ICONCLASS concepts).

The details of the HW can be found here: 

Decisions: 

for point (i) 

HW: GH & MD to collaborate on proposing the terms and hierarchies for the categories specializing type of E4 Period and E92 STV identified 
HW: GH, MD, AK, AG can try and map the terms for place types found in Alexandria, TGN & Geonames  

for point (ii)

HW: TV to consolidate these items to the list, also to draft a text that states what has been done and the resources that have been used in the process.

Marseille, October 2023

HW by Athina Kritsotaki (private communication -- 21 February 2024) 

Dear all,

I am sending you my part: I used Martin's excel and I added a list of terms I found from TGN and Geonames (geonames full list). I also tried a
first mapping. Additionally, I attach an explanatory text for place types from TGN (that explains the partial list that is available on the Getty site)

Best,
Athina

In the 58th CIDOC CRM & 51st FRBR/LRMoo SIG Meeting, the SIG reviewed the HW ([1] and [2]) submitted by AK, where she provided type-restrictions for instances of E4 Period. For a summary of the points raised see here. 

How to move forward – HW assignment: 

• Available definitions from the TGN to be consulted
• GH & OE will work towards providing a definition of the different types of space-time phenomena used to restrict place types  

Nb. No progress has been made wrt consolidating (i) the functional role of minimal vocabularies, (ii) terms useful for migration paths, and (iii) type-restrictions for classes and typed properties in a guidelines document. 

Paris, March 2024
 

In the 59th CIDOC CRM SIG & 52nd FRBR/LRMoo SIG joint Meeting, participants reviewed the proposal 
 

A. to provide a list of recommendations

1. (i) (concerning deprecated CIDOC CRM class E40 Legal Body) from AAT –besides corporations, and corporate bodies –namely, advise to use relevant/appropriate terms listed under
   1. AAT ID: 300024980 <people by occupation>
   2. AAT ID: 300025825 <people by activity>
2. (ii) (concerning the deprecated CIDOC CRM class E82 Actor Appellation) from AAT –besides personal names –toponyms and also actor types
   1. AAT: ID: 300404655 place names
   2. AAT: ID: 300025825 <people by activity>
3. (iii) (concerning deprecated CIDOC CRM class E46 Section Definition) from BBT –namely, advise to use terms
4. BBT ID: 0000049 geometric extents
5. BBT ID: 0000050 points
6. BBT ID: 0000053 linear extents
7. BBT ID: 0000051 surface areas
8. BBT ID: 0000052 d-volumes

B. to provide a definition for “preferred identifier assignment” (in order to specify E15 Identifier Assignment):

• CRM Thesaurus term: 
  Scope note: the process of assigning an identifier which is to be preferred for use within a context, such as an organisation or a project.

Decisions: 
For point (A): 
- The SIG accepted the proposal for item (ii): type restrictions from BBT for deprecated CRM Class E46 Section Definition –the scope notes for the relevant set of terms can be found here. 

- The SIG rejected the proposal for items (i) and (ii): type restrictions from AAT relating to <a> ID: 300024980 <people by occupation>, and <b> ID: 300025825 <people by activity> for deprecated classes E40 and E82 –they didn’t see the relevance and they claimed the proposed terms were a far worse fit than the original proposals –namely: ID: 300025969 corporations & ID: 300386361 corporate bodies for E40, and ID: 300266386 personal names for E82.

- -  Coroporate bodies are not people, so then people by activity and people by occupation are not groups.

- - The AAT hierarchies are also not relevant for E82. 
 

For point (B):

- The motivation to create a term for “preferred identifier assignment” comes from the fact that the scope note of E15 explicitly mentions that one can express the concept better through assigning a type to E15 via P2 has type: E55 Type (“preferred identifier assignment”). 

- The SIG was not OK with that, instead they suggested that we do not make a recommendation and that this clause in the scope note of E15 is deleted. To be dealt with through a new issue. 

- -[…] The fact that an identifier is a preferred one for an organisation can be expressed by using the property E1 CRM Entity. P48 has preferred identifier (is preferred identifier of): E42 Identifier. It can better be expressed in a context independent form by assigning a suitable E55 Type, such as “preferred identifier assignment”, to the respective instance of E15 Identifier Assignment via the P2 has type property.
 

Post by George Bruseker (18 November 2024)

Dear all,

I guess I just wanted to draw attention to the fact that multiple groups
seem to be trying to do the same thing and dialogue and sharing of
knowledge and approaches might always be a useful approach.

In particular there are LIDO versions of this exercise here:

https://terminology-view.lido-schema.org/vocnet/?startNode=lido00409&la…

and Linked.Art versions of the same exercise here:

https://linked.art/model/vocab/

Given that there is quite some progress in this activity in these groups,
it could be useful to consult what has already been done ground up.

Best,

George