Issue 484: 7.0 preparation - missing examples

ID: 
484
Starting Date: 
2020-02-28
Working Group: 
4
Status: 
Open
Background: 

Posted by Chryssoula on 29/4/2020

Following the decisions of last sig (46th joint meeting of the CIDOC CRM SIG and ISO/TC46/SC4/WG9; 39th FRBR - CIDOC CRM Harmonization meeting) issue 483. New issue created to collect the hw for the missing examples.

Heraklion, 29/4/2020

 

Current Proposal: 

Posted by Steve on 28/2/2020

Here are Pat’s examples for P130.

Rgds

SdS

 

Stephen Stead

Tel +44 20 8668 3075

Mob +44 7802 755 013

E-mail steads@paveprime.com

LinkedIn Profile https://www.linkedin.com/in/steads/

---------------------------------------------------

From: Pat Riva <pat.riva@concordia.ca>
Sent: 28 February 2020 13:54
To: Stephen Stead <steads@paveprime.com>
Subject: Examples for P130

Hi Steve,

Some examples to choose from, not currently used in LRMoo or CRMbase. The scope note also indicates this could be used for more archaeological situations, I leave those to you.

Pat

Posted by Steve on 28/2/2020

Hi Martin

I have created 2 examples one for a spacetime volume derived from an event and one derived from a physical object. I have added the extra step to time primitive to illustrate the difference between STV, Time-Span and Time Primitive in the hope that this will make things clearer to the user community. I welcome your comments on the approach.

·      The Spacetime Volume of the Battle of Waterloo 1815 P160 has temporal projection The Time-Span of the Battle of Waterloo [P82 at some time within Sunday, 18 June 1815 (E61 Time Primitive)]

·      the spatio-temporal trajectory of the H.M.S. Temeraire from its building in 1798 to its destruction in 1838 P160 has temporal projection The Time-Span of the existence of H.M.S. Temeraire [P82 at some time within 1798-1838 (E61 Time Primitive)]

Posted by  Steve on 1/3/2020

Please find attached another piece of the property example homework

Posted by Martin on 3/3/2020

On 2/28/2020 4:50 PM, Stephen Stead wrote:
>
> Hi Martin
>
> I have created 2 examples one for a spacetime volume derived from an event and one derived from a physical object. I have added the extra step to time primitive to illustrate the difference between STV, Time-Span and Time Primitive in the hope that this will make things clearer to the user community. I welcome your comments on the approach.
>
> ·      The Spacetime Volume of the Battle of Waterloo 1815 P160 has temporal projection The Time-Span of the Battle of Waterloo [P82 at some time within Sunday, 18 June 1815 (E61 Time Primitive)]
>
> ·      the spatio-temporal trajectory of the H.M.S. Temeraire from its building in 1798 to its destruction in 1838 P160 has temporal projection The Time-Span of the existence of H.M.S. Temeraire [P82 at some time within 1798-1838 (E61 Time Primitive)]

The second is a good one, because E18 is no more an STV. The Battle of Waterloo is an E7, and hence it is not recommended to use P160, but P4, once from E4 downwards P4 = P160.

So,

I propose

"The Battle of Waterloo 1815 (E7) P4 has time-span (= P160 has temporal projection) The Time-Span of the Battle of Waterloo [P82 at some time within Sunday, 18 June 1815 (E61 Time Primitive)]

or drop it.

We may better use a declarative spacetime volume, such as "Great Britain in the 19th century" ... 1800-1900

?

Posted by Martin on 4/3/2020

Hi Steve,

The point is that P160 is implied by many other properties, and more for querying than declaring. So, may be even one example is good enough.

Interesting are physical things with no known genesis event:

" the spatio-temporal trajectory of Allende Meteorite fallen on Earth February 8, 1969" P160 has temporal projection The Time-Span of the existence of the Allende Meteorite [P83 ongoing throughout 4.000.000.000 BC-2020AD (E61 Time Primitive)]

(https://www.smithsonianmag.com/smithsonian-institution/oldest-material-s...)

Posted by Athina on 5/3/2020

E96 Purchase

the purchase of 10 okka of nails by the captain A. Syrmas on 18/9/1895 in Thessaloniki.

P179 had sales price (was sales price of)

the purchase of 10 okka of nails by the captain A. Syrmas on 18/9/1895 (E96) had sales price 20 piastre (grosi) (E97).

The specific examples were used from “Account Book of ship D.S.Skyliytsis 28/9/1895-19/10/1896 Captain A.Syrmas,Book no5; Inventory Number A.E 53/01, E.L.I.A, Athens” in the context of SEALIT project (http://www.sealitproject.eu/)

Posted by Martin on 9/3/2020

The centroid from https://sws.geonames.org/735927 (E53) P168 place is defined by [40°31'17.9"N 21°15'48.3"E] (E94) (a single point for approximating the centre of the city of Kastoria, Greece)

Martin’s coordinates for Kastoria (E53) P168 place is defined by [40°30'23"N 21°14'53"E, 40°31'40"N 21°16'43"E] (E94) (a square covering the built settlement structure of Kastoria, Greece)

Martin’s centroid for Kastoria (E53) P168 place is defined by [40°30'55"N 21°15'47"E] (a point in the lake of Kastoria in the centre of the area covered by the city)

Posted by Martin on 9/3/2020

Hi Steve, what about these examples: 
P156 occupies (is occupied by)

Domain:             E18 Physical Thing

Range:                E53 Place

 

Subproperty of:    E18 Physical Thing. P157i (provides reference space for) :E53 Place

Subproperty of:    E18 Physical Thing:P53 has former or current location (is former or current location of): E53 Place

 

Quantification: one to one (0,1:1,1)

 

Scope note:  This property describes the largest volume in space, an instance of E53 Place, that an instance of E18 Physical Thing has occupied at any time during its existence, with respect to the reference space relative to the physical thing itself. This allows for describing the thing itself as a place that may contain other things, such as a box that may contain coins. In other words, it is the volume that contains all the points which the thing has covered at some time during its existence. The reference space for the associated place must be the one that is permanently at rest (P157 is at rest relative to) relative to the physical thing. For instances of E19 Physical Objects it is the one which is at rest relative to the object itself, i.e. which moves together with the object. For instances of E26 Physical Feature it is one which is at rest relative to the physical feature itself and the surrounding matter immediately connected to it. Therefore there is a 1:1 relation between the instance E18 Physical Thing and the instance of E53 Place it occupies. We include in the occupied space the space filled by the matter of the physical thing and all its inner spaces.

 

This property implies the fully developed path from E18 Physical Thing through P196 defines, E92 Spacetime Volume, P161 has spatial projection, E53 Place. However, in contrast to P156 occupies, the property P161 has spatial projection does not constrain the reference space of the referred instance of E53 Place.

 

In contrast to P156 occupies, for the property P53 has former or current location the following holds:

·        It does not constrain the reference space of the referred instance of E53 Place.

·        It identifies a possibly wider instance of E53 Place at which a thing is or has been for some unspecified time span.

·        If the reference space of the referred instance of E53 Place is not at rest with respect to the physical thing found there, the physical thing may move away after some time to another place and/or may have been at some other place before. The same holds for the fully developed path from E18 Physical Thing through Pxxx defines, E92 Spacetime Volume, P161 has spatial projection, E53 Place.

 

In First Order Logic:

P156(x,y) ⊃ E53(y)

                           P156(x,y) ⊃ E18(x)

P156 (x,y) = [E18(x) ∧ E53(y) ∧ P196(x,z) ∧ P161(z,y) ∧ P157(y,x)]

Example:

Burg Eltz (english: Eltz Castle) near Koblenz, Germany P156 occupies the space within the outer walls of Burg Eltz since 1661AD (E53) (the castle has been extended from the 12th century until 1661AD and not been destroyed up to present, containing buildings from various periods)

The Saint Titus reliquary P156 occupies the space of the Saint Titus reliquary (the reliquary is currently kept in the Saint Titus Church in Heraklion, Crete since 1966 and contains the skull of Saint Titus)

Posted by Steve on 10/3/2020

May I suggest the following minor refinement

Martin’s centroid for Kastoria (E53) P168 place is defined by [40°30'55"N 21°15'47"E] (a point in the lake of Kastoria at the geographic centre of the area covered by the city)

Also could/should we make Martin’s centroid the exact centre of Martin’s square?

Posted by Martin on 11/3/2020

On 3/10/2020 12:08 AM, Stephen Stead wrote:
>
> May I suggest the following minor refinement
>
> Martin’s centroid for Kastoria (E53) P168 place is defined by [40°30'55"N 21°15'47"E] (a point in the lake of Kastoria at the geographic centre of the area covered by the city)
>
> Also could/should we make Martin’s centroid the exact centre of Martin’s square?

Here it is:

Martin’s centroid for Kastoria (E53) P168 place is defined by [40°31'01.5"N 21°15'48"E] (a point in the lake of Kastoria in the centre of the area covered by the city)

Alexander v. Humboldt's measurement for the Plaza Mayor in Cumaná, Sucre,Venezuela 1799-1800AD (E53) P168 place is defined by [10°27'52"N 66°30'02"W] (actually 260km west of Cumaná)

Citation: Humboldt, Alexander von : Die Südamerika-Reise (the only German edition authorised by A. v. Humboldt): Original titel: Reise in die Äquinoktial-Gegenden des Neuen Kontinents (German Edition) . eClassica. Kindle Edition. 2015

and P189:
Alexander v. Humboldt's measurement for the Plaza Mayor in Cumaná, Sucre,Venezuela 1799-1800AD (E53) P189 approximates the Plaza Mayor in Cumaná (probably today's Plaza Andrés Eloy Blanco. Von Humboldt's measurement had an error of 260km to the west).

Posted by Martin on 11/3/2020

P170 defines time (time is defined by)

Domain: E61Time Primitive

Range: E52 Time Span

Scope note:         This property associates an instance of E61 Time Primitive with the instance of E52 Time Span it defines.

In First Order Logic:

                           P170(x,y) ⊃ E61(x)

                           P170(x,y) ⊃ E52(y)

 Examples:

(1800/1/1 0:00:00 – 1899/31/12 23:59:59)(E61) P170 defines time The 19th century (E52)

(1968/1/1 – 2018/1/1)(E61) P170 defines time “1968/1/1 – 2018/1/1” (E52) (an arbitrary smaller time-span during which the Saint Titus reliquary was present in the Saint Titus Church in Heraklion, Crete) 

Posted by Martin on 11/3/2020

Dear All,

Here my example (geojson, gml, xsd)  for P169. Gerald, is there a good syntax in geosparql or WKT?

Gerald, do have an example of an expanding city, that could easily be approximated by two STV boxes?

Posted by Martin on 12/3/2020

Dear All,

I kindly ask for help to produce more nice examples for the new temporal primitives. This should be the domain of historians and archaeologists.

I found so far the following:

P173:

The legendary run from Marathon to Athens 490BC (E7) P173 starts before or with the end of  The Battle of Marathon 490BC (E7)

P183:

Gisle taking office as Bishop of Linköping 1139(E7) P183 ends before the start of The Guta saga composition (E65)

See “…the account of the church-building and the arrangements concerning the Bishop of Linköping advance that by a further century or so, since the first recorded bishop was Gisle, who took office in 1139. This gives a terminus postquem of circa 1140 for the saga’s composition.”

In:
Title      Guta Lag and Guta Saga: The Law and History of the Gotlanders
Routledge Medieval Translations
Editor    Christine Peel
Edition  illustrated
Publisher            Routledge, 2015
ISBN      1317565258, 9781317565253

The complete list is the following

  1. P173 starts before or with the end of
    • Astart ≤ Bend
  2. P174 starts before the end of
    • Astart < Bend
  3. P175 starts before or with the start of
    • Astart ≤ Bstart
  4. P176 starts before the start of
    • Astart < Bstart
  5. P182 ends before or with the start of
    • Aend ≤ Bstart
  6. P183 ends before the start of
    • Aend < Bstart
  7. P184 ends before or with the end of
    • Aend ≤ Bend
  8. P185 ends before the end of
    • Aend < Bend

Posted by Gerald on 27/3/2020

Dear Martin and all,

In GML (http://www.ogc.org/standards/gml) it is possible to express temporal geometries and GeoSPARQL has the property “asGML".
I created a GML example of Innsbruck with a geometry for 1805 (2D to keep it easy) and another geometry for 2020. 
The GML was created using QGIS and uses other xml tags (ogr: //ogr.maptools.org/) from GDAL (https://gdal.org/) that make it possible to load and display the data in QGIS. 
I added the original GML tag for gml:TimeInstant to by compliant with the gml specification. It was not present in the original export from QGIS.
That’s one of the problems with GML specifications. Not everything is implemented in software packages. 

Geojson became an OGC standard in 2019 (http://www.ogc.org/standards/eo-geojson) and has possibilities to encode temporal geometries within its properties, but that is no more than attributes.

We could also look into the OGC “Moving features” which seem very similar to our Spacetime Volumes: 
OGC® Moving Features
https://www.ogc.org/standards/movingfeatures/

There is an XML and a csv standard for encoding moving features, but I believe the extension for geojson could be the most promising: 
"OGC Moving Features Encoding Extension – JSON"

https://docs.opengeospatial.org/bp/16-140r1/16-140r1.html

Example:

{

    "type": "MovingPoint",

    "coordinates": [ [100.0, 0.0], [101.0, 0.0], [101.0, 1.0]],

    "datetimes": ["2011-07-14T22:01:01Z", "2011-07-14T23:01:01Z", "2011-07-15T00:01:01Z"],

    "interpolations": ["Linear"] // an interpolation method during the period of datetimes

}

I hope this helped,
 

Posted by Martin on 28/3/2020

Dear Christian-Emil, George

attached my complete list of examples.

Christian-Emil, please try to find the missing examples.

George, please prepare an e-mail vote for the given ones.

Posted by Christian Emil on 31/3/2020

Dear all,

Here is an example of a terminus post quem and teminus ante quem taken form Regesta Norvegica (somewhat complicated)

Best,

Christian-Emil

Regesta number: 601

Date [1422 – before 1426 May 20?] ¹

Vidimus [confirmed copy] by Ture Karlsson, Karl Ulvsson, Eivind Torgilsson, Gyrd Toresson, Tjodolv Gyrdsson and Tord Toraldsson: Transcript of RN II no. 1075 [DNVI No. 374].

Printed: DN VI No. 483 (incorrectly dated [around 1420-1440]).

¹ Term[inus] p[ost] q[uem] is based on the fact that the issuer Ture Karlsson first appeared in Norway in 1422 (cf. no. 193 earlier).

Term[inus] a[nte] q[uem] is based on the following reasoning: 

Issuer Karl Ulvsson, Catherine Ulvsdaughter's brother, is first mentioned in 1410 (RN IX no. 822). Catherine Ulvsdaughter confirms her mother's gift of Forsvik mill [to Vadstena Convent in Sweden] in a letter dated May 20, 1426 (cf regesta no. 602 later) . Her husband, Herlaug Petersson,   makes a confirmed copy [vidius] date July 11, 1426  of a letter her father Ulv Holmgeirsson's received in 1364 (RN VI no. 1034).[Karl Ulvsson’s and Catherine Ulvsdaughter's father received  this  letter from the king confirming that he got Forsvik​ mill back from some third person]  This indicates that Karl Ulvsson died sometime before May 20, 1426. Otherwise normally he, as heir, would have been responsible for both the letter of confirmation [vidimus]  and the testimony of the gift.

Posted by Martin on 31/3/2020

Dear Christian-Emil,

That is a great example for CRMInf by the way! "Otherwise normally he, as heir, would have been responsible for both the letter of confirmation [vidimus]  and the testimony of the gift" and temporal order of causal relations is the complete inference logic.

Please explain DN VI No. 483: What does this document pertain to? Who is the issuer, an attorney, a scribe, or a private person?

I'll try to turn that into examples.

We are not only looking for termini postquem and antequem. Please check all the remaining temporal relations, what could demonstrate them. I think about declarative statements about archaeological periods.

Posted by George on 4/1/2020

Dear all,

Here is a possible format for the e-vote. Do you find it intuitive enough? I think that we need to have them input their email addresses because we can’t have just anybody voting…

https://docs.google.com/forms/d/e/1FAIpQLSf0_502Sxk0gpoRC58YnT18Vp8E2qbH6skBiG1yboCEDplpTQ/viewform

Posted by Martin on 1/4/2020

Dear George,

The form is VERY good.

I propose choices: YES NO VETO. NO and VETO with comment field. Please mark property labels in the example with italics. Pxxx is not for example vote. It should go into a different form, and should be voted "overall", "scope note", "example".

 

Posted by George on 2/4/2020

Dear Martin et al.,

Ok now the examples are in one form and the new property proposal is another form.

Examples:

https://docs.google.com/forms/d/e/1FAIpQLSf0_502Sxk0gpoRC58YnT18Vp8E2qbH6skBiG1yboCEDplpTQ/viewform

New Property:

https://docs.google.com/forms/d/e/1FAIpQLSceFwCsX8HnsO551sSXh4Iol0wRLCyDyxFsbRD-mp-wuy21FA/viewform

I imagine we don't want to create endless vote forms, so do we have an idea of the things that nee to be voted for v 7.0? I think we should do all of them at the same time in the least amounts of forms possible so people don't get lost.

Other thoughts: it is not really possible to have an option, yes, no, veto and other with google forms. We will have to follow the logic that Eleni suggested which is pretty much the same anyhow. Either clearly vote yes or no, if you have a comment then choose the comment option and write whatever caveats you like. The most extreme caveat is veto. Since if you veto you should say why, this seems sensible.

For the new property question, I had to vary the form some. I think it still makes sense though it is a touch repetitive. 

Another good suggestion Eleni made was that it would be good to only email the actual voting members with these forms rather than just anyone who is on the SIG list (they are not all members or membership has little meaning). Do we have a mailing list of just the actual members of SIG?

 

Posted by Steve on 3/4/2020

Hi George

The examples form allows only one completion (which is good!). However, the New Property form has no such restrictions.

I would vote for several forms so that the work of reviewing and voting can be divided up into smaller sessions. I found that the examples took me several sittings and I was constantly worried that I would lose all my input or accidentally commit and not be able to go back (as it is one completion only form)

Posted by Thanasis on 6/4/2020

One can split Google Forms is separate pages, where you need to click "next" to move to the following question, but I have just done an input persistency test and the answers do not seem to remain in the form if the session ends. So although it will make the process more manageable in terms of the size of the form, it might create a false sense of security regarding persistence of data.

Posted by Thanasis on 6/4/2020

Maybe group them by relevance? Maybe something like this:

E93 Presence
P195 was a presence of (had presence)
P169 defines spacetime volume (spacetime volume is defined by)
P166 was a presence of (had presence)

P160 has temporal projection (is temporal projection of)
P164 during (was time-span of)
P170 defines time (time is defined by)
P173 starts before or with the end of (ends after or with the start of)
P183 ends before the start of (starts after the end of)

P168 place is defined by (defines place)
P167 at (was place of)
P156 occupies (is occupied by)

P130 shows features of (features are also found on)
E96 Purchase
P179 had sales price (was sales price of)

Unless there was already a logical sequence which I missed.
 

Posted by Thanasis on 20/4/2020

Dear all,

Having received some notes from Chryssoula I went through the examples of CRMbase and CRMsci and I picked out what I could identify as consistent practices. I have put these in the attached documents and in some cases I have made some proposals to handle some special cases a bit better. Happy to discuss in the next meeting. I hope these templates do not contradict any current practices.

I also noticed that there are a lot of inconsistencies with the examples and we will need to scan through the document to fix these before the release.

Some issues which I picked for example are:
E55 Type: The examples are lists of types, not plain types.
E57 Material: Capitalisation is inconsistent with the rest of the document.

==================================================

CB's note: The discussion about example templates continues on issue 493

Following votes were called on April 16th, 2020 with a deadline of April 30th, 2020, The CRM SIG editorial team met virtually to evaluate the results and put them into action. The added examples belong to the following categories and can be found here

1 Miscellaneous Examples

2 Space Time Volume and Temporal Properties Examples

3 Presence Class and Related Properties Examples

Generally the votes had to do with the addition of examples to properties and classes in the standard in view of the upcoming release of version 7.0 of CIDOC CRM. 

30 April 2020