CRMarcheo generalization of the properties AP12 confines and AP11 has physical relation

Posted by Christian Emil on 6/3/2019

Dear all,
In the Norwegian digital excavation records I have inspected, one records the information that a profile (instance of A10 Excavation Interface) cuts through one or many stratigraphic structures (instances of A8 Strategraphic Unit). This is common in a non-single-context excavation, where vertical profiles are dug through structures, say postholes. This information is currently not possible to express in a CRMarcheo compliant form.  In the text below I argue for a small amendment of CRMarcheo. The proposal is to drop the highly specialized A12 Confines and generalize A12 has physical relation to a property of CRMsci. I have discussed this issue briefly with Gerald Hiebel and Achille Felicetti. ​ All errors are mine.

I attach a small ppt.

Best,
Christian-Emil

​Background

****Relevant issues

Issue 283: Add superproperties to properties of CRMarcheo
Issue 337: Excavation Interface

****Earlier discussions:

Outcome of the Berlin meeting (42nd) November 2018:

- the new A10 Excavation Interface class to be included in CRMarchaeo documentation, also there is no need to introduce a new APxx property since O7 confined could be used instead. O7 has as range S10 Material Substantial but this is consistent with S22 Segment of Matter, as once is dug out it corresponds to an S10.

- Adjustment of the range of the property AP4 Created Surface from S20 Rigid Physical Feature  to A10 Excavation Interface

Discussion in the Cologne meeting (40th) January 2018):

The SIG expressed concerns about to many confine properties and suggested that one could introduce a common superclass of the two classes A10 Excavation Interface and A3 Stratigraphic Interface, and also a superclass for the A2 Stratigraphic Volume Unit.

Response to this discussion in the Cologne meeting

In both geological and archaeological stratigraphy, we find the concepts surface/interface and the layer/stratum. A surface/interface has no volume in contrast to a layer/stratum.  The following classes and the connected properties can be used to model geological stratification as well as archaeological:

A2 Stratigraphic Volume Unit

A3 Stratigraphic Interface

A4 Stratigraphic Genesis

A5 Stratigraphic Modification

A7 Embedding

A8 Stratigraphic Unit

The three classes below (and the connected properties) involve actors (humans) and could be used in geological fieldwork and definitely in paleontological fieldwork:

A1 Excavation Process Unit

A6 Group Declaration Event

A9 Archaeological Excavation

The CRMarceo is to a large extent well suited to model geological stratigraphy as well as archaeological.  If there had been a demand for an extension for geology, one should considere a common stratigraphic extension “CRMstrat” for a “CRMgeology” and  CRMarcheo. Since this is not the case, one should stick to CRMarcheo as an extension of CRMsci.  Still CRMarcheo can be simplified by “lifting” some of the properties from CRMarcheo into CRMsci.

For example, in the Norwegian digital excavation records I have inspected, one records the information that a profile (instance of A10 Excavation Interface) cuts through one or many stratigraphic structures (instances of A8 Strategraphic Unit). This is common in a non-single-context excavation, where vertical profiles are dug through structures, say postholes. This information is currently not possible to express in a CRMarcheo compliant form.  There are at least two solutions:

Solution 1

Extend CRMarcheo with two properties similar to AP11 has physical relation (is physical relation of):

APxx  has physical relation (is physical relation of):

Domain:A10 Excavation Interface

Range: A8 Stratigraphic Unit

APyy  has physical relation (is physical relation of):

Domain:A10 Excavation Interface

Range: A10 Excavation Interface

In addition one will need a new property

Ayy confines (is confined by)

Domain: A10 Excavation Interface

Range:A2 Strategraphic Unit

Solution 2

Lift (generalize)   the domain and range AP11 has physical relation (is physical relation of) to ‘S20 Rigid Physical Feature’ and either keep it as a property of CRMarcheo or better make it a new property in CRMsci. The property AP11.1 has type will be replaced by a .1 for the new property.

Use  the CRMsci property O7 confined (was confined by)  everywhere and delete AP12 confines (is confined by) 

To avoid this proliferation of properties , the following proposal is made:

Proposals:

1)      Delete AP12 confines (is confined by)  and use O7 confined (was confined by)

2)      Lift (generalize) the domain and range of AP11 has physical relation (is physical relation of) to ‘S20 Rigid Physical Feature’ make it a new property in CRMsci.

Oxx has physical relation (is physical relation of)

Domain: S20 Rigid Physical Feature

Range: S20 Rigid Physical Feature

The property AP11.1 has type will be replaced by a Oxx.1 for the new property.

3)   

The name of AP11 is misleading since it indicate that the range instance is the type of relation and not another stratigraphic layer/surface which is in relation to the domain instance​.  The cardinality should be (0,n, 0,n) and not (0,n:0,1).   A draft of the new property can be

APxx  is in physical relation to

Domain:
S20 Rigid Physical Feature
Range: S20 Rigid Physical Feature

Quantification: many to many (0,n:0,n)

Scope note: This property identifies the physical relationship between two
S20 Rigid Physical Feature. The type of physical relationships is documented through the property  Oxx.1 has type

Properties: Oxx.1 has type: E55 Type

Example can be taken from archaeology and geology

​

​