Issue 521: cardinality corrections with E93 Presence.
Posted by Martin on 30/10/2020
Dear All
P195 was a presence of (had presence)
Domain: E93 Presence
Range: E18 Physical Thing
Quantification: many to one, necessary (1,1 : 0,n)
Is wrong. This is a shortcut through P166, which is necessary, but an instance of E93 Presence may be defined on spacetime volumes others than that of an E18.
Therefore:
NEW Quantification: many to one (0,1 : 0,n)
Further:
P197 covered parts of (was partially covered by)
Domain: E93 Presence
Range: E53 Place
Scope note:
This property associates an instance of E93 Presence with an instance of E53 Place that geometrically overlaps with the spatial projection of the respective instance of E93 Presence. A use case of this property is to state through which places an object or an instance of E21 Person has or was moved within a given time-span. It may also be used to describe a partial or complete, temporary or permanent extension of the spatial extent of some realm into a neighboring region during a known time-span. It may also be used to describe a partial or complete, temporary or permanent extension of the spatial extent of some realm into a neighboring region during a known time-span. It is a shortcut of the more fully developed path from E93 Presence through P161 has spatial projection, E53 Place, P121 overlaps with to E53 Place.
....needs a Quantification: many to many (0:n,0:n)
An instance of E53 Presence may cover parts of many places, and places may not be covered by any presence or by many.
One may argue, that the spatial projection of an instance of E93 Presence is a subproperty of P197, if P121 would include identity. I vote against such an interpretation, and for proper "overlaps".
I propose an immediate e-vote to be called for this issue.
Posted by CEO on 1/11/2020
Dear all,
This is in fact three issues
1) The correction of the quantification of P195
2) Missing quantification for "P197 covered parts of "
3) Does" P121 overlaps with" cover identity?
1 and 2 can be easily be solved as Martin indicates.
Number 3 is more complex:
In general: Should the part of properties be non reflexive?
P7 took place at (witnessed)
Domain: E4 Period
Range: E53 Place
Scope note: [...] It is a shortcut of the more fully developed path from E4 Period through P161 has spatial projection, E53 Place, P89 falls within to E53 Place
If A P161 B, that is, P is the spatial projection of A, then most of us will accept A P7 B which implies that B P89 B. Therefore P89 is reflexive and includes identity. It is confusing (an inconsistent) if P121 cannot include total overlap.
The issue has been resolved by the work of the editorial team.
- The cardinality of P195 has been set to many to one, necessary (1,1 : 0,n) in CIDOC CRM 7.1.1 (official)
- The cardinality of P197 has been set to many to many (0,n : 0,n) in CIDOC CRM 7.1.1 (official)
- All properties with part of semantics have been declared non-reflexive (P5, P46, P106) in CIDOC CRM 7.1.1 (official)
The issue is closed
May, 2021