Issue 267: CRMarchaeo - Congratulations and further comments

ID:

267

Starting Date:

2014-09-25

Status:

Done

Closing Date:

2015-02-12

Background:

Posted by Keith May, Paul Cripps, Doug Tudhope, Ceri Bindingon on 25/9/2014

Dear CRM-SIG,

We understand that the latest version 1.2 of CRMarchaeo (http://www.ics.forth.gr/isl/index_main.php?l=e&c=711) is due to be presented at the SIG meeting next week. This is welcome news and very good progress.

However despite some of us having been closely involved in the early 'birth & development' of these extensions We are not sure at this point what the latest trajectory for comments on CRMarchaeo is?

Is there a mechanism whereby the current version will be made open for 'Review' to CRM-SIG members who cannot attend the meeting, either before, after, or as part of the agenda of the meeting and if so can you provide direction on that?

The last version we saw (after CAA Paris) dated April2014 looked very positive and definitely reflected major progress and excellent efforts towards completing this complex work. However it remained unclear at that time (and still) how some of our feedback on more detailed modelling of the stratigraphic and physical relationships that derived from CRM-EH modelling (http://www.cidoc-crm.org/docs/Ontological_Modelling_Project_Report_Sep2004.pdf) and further research potential arising from implementations of such modelling during the STAR project (http://hypermedia.research.southwales.ac.uk/kos/star/) had been represented in the drafts of CRMarchaeo. We’d particularly like to see a bit more detail on the physical relations (such as Filled by, Fill of, Cut by, Cuts) between A8 Stratigraphic Unit and A2 Stratigraphic Deposit Unit and A3 Stratigraphic Interfaces, along with more consideration of the stratigraphic inferencing that is proposed for A5 Stratigraphic Modification using AP13 'Has stratigraphic relation'. The suggestion for an appendix that "illustrates an example for stratigraphic and spit excavation and shows how these two types of excavation methodology would be represented with CRMarchaeo" sounds like a very helpful idea, but I don't know if that has been achieved yet?

Also we note that the reference document for CRMarchaeo still has a number of scope notes to be completed (e.g. A5 Stratigraphic Modification) which makes consideration of some of these issues harder.

In Summary,

There are no dramatic incompatibilities between CRMarchaeo as it now stands and CRM-EH. The further modelling of CRM-EH took a slightly different path and clarified the CRM-EH classes, subtyping them where necessary to achieve a more explicit model, but one which is geared to representing the single context recording methodology used in the UK (and more widely). The more generic approach taken in CRMarchaeo is to accommodate a broader range of excavation techniques (Spits, Planum, Locus, etc).

For the enhancement of CRMarchaeo we can provide more detailed feedback on the following if directed how.

Some clarification on the relationship between A2 and A3 is needed, particularly wrt AP12. Likewise with the A2 genesis and contains/confines. Also more could be done to represent the temporal relationships for the events leading to the stratigraphic sequence/matrix which is so integral to relating much of the other data from excavations. We think from STAR research that some sub-properties of Allen, to accommodate stratigraphic sequences, may be required and prove very beneficial for integration and inferencing over stratigraphically related datasets (e.g. a 'Stratigraphically directly before/after' is not necessarily temporally synonymous with the Allen p120 Occurs Before/Occurs After ).

It would be advantageous if such small but very significant amendments could be incorporated in the CRMarchaeo model rather than requiring further extensions in the future.

So the main question we have is how best to feed these elements into the process for acceptance and use of CRMarchaeo?

In 31st joined meeting of the CIDOC CRM SIG, ISO/TC46/SC4/WG9 and the 24th FRBR - CIDOC CRM, discussing the CRMarcheo made the following comments:

· Achille and Paola will improve the scope notes.

· The status of the document is a draft under revision by Gerald Hiebel . (new version by Gerald is attached to this issue)

· Gerald will communicate with EH-CRM Keith May.

· Carlo and Gerald will work together on how to represent AP14

Heraklion, Crete, October 2014

Posted by Gerald Hiebel on 21/10/2014

Dear Keith,

Thanks for the message and sorry for the late reply from my side. Chryssoula is preparing the latest version of CRMarchaeo for the CRM-SIG minutes and will send it around with the tasks that have been assigned and discussed.

As I just finished the parts that have been discussed in the CRM-SIG meeting, I send you preliminary this latest version as I do not know when Chryssoula will be able to send out the minutes and I wanted to do it for a long time. Please have a look at the proposed theoretical solution for relating stratigraphic to physical relationships. How to implement it is not clear to me yet, but I am not a guru there.

What we would need is a paragraph on EHCRM compatibility. Maybe you could make a proposition or we work together on something? Still scope notes have to be improved....

Best,

Gerald

Posted by Keith May on 29/10/2014

Hi Gerald,

I’m happy to work with you on a piece of text for the CRMarchaeo documentation to cover CRM-EH/CRMarchaeo compatibility – indeed it’s something I think we’ll need to advertise in future presentations and in papers more widely as the CRMarchaeo comes into wider circulation and use and people will want to know how it fits with earlier CRM-EH work.

I like the inclusion in the document of the stratigraphic diagrams which certainly help to explain the fundamentals for archaeologists (and even with the ‘spit’ inserted to refer to differing methodologies! J ).

We briefly mentioned in the previous email to CRM-SIG that we don’t see any ‘dramatic incompatibilities between CRMarchaeo as it now stands and CRM-EH“ .

But how we phrase the ‘degree of compatibility’ in the text for the document, for me and Paul, rather depends on the degree to which the comments about the stratigraphic relations and reasoning sent to CRM-SIG are taken on board.

I cannot see in the text document you have sent that it has accounted for some of the issues we highlighted with how Cuts, Fills and Interfaces actually relate.

While an A3 Interface (of E55 type ‘cut’) can/will confine an A2 deposit (of E55 type ‘Fill’)

A3 Interfaces (of type cut) always also ‘cut’ other A2 deposits which they do not AP12 confine

e.g. In the section example Fig 4. A3 [3] does ‘AP12 confine’ but it also physically ‘cuts’ (4) & (15) without AP12 confining them at all.

I still don’t think this works as currently shown in the modelling inserted in the document (and I think is not helped by “AP5 Cuts” relation being used for the archaeological activity of A1 Excavation Process Unit excavates an A8 Stratigraphic Unit).

I also note a highlight in the scope note text of A3 which may relate to some uncertainty on this.

One event of removal process may produce many stratigraphic interfaces - I wouldn’t say so. It produces One stratigraphic interface, and Many physical interfaces.

1 stratigraphic relation = 1 event (of uncertain duration). I think that’s pretty fundamental to stratigraphic recording.

I do think some of this might be helped if we see what you mean by the outcome of “Chryssoula is preparing the latest version of CRMarchaeo for the CRM-SIG minutes and will send it around with the tasks that have been assigned and discussed”

Paul wrote some fuller comments which I only outlined in my email to the CRM-SIG for the sake of brevity, and because I was hoping for a response that suggested who we should actually address them to. I am sending these now to you, Maria and Martin (see attached for ref), as they may help explain more of the specific issues and how we think they might be resolved.

The CRMarchaeo is already very good, but I think it would be great if we can just get these small but important revisions/additions included, which are also supported from Doug and Ceri’s actual implementation experience in STAR, and which I think will raise issues if ARIADNE plans to implement CRMarchaeo using actual archaeological data.

Posted from Gerald Hiebel on 10/11/2014

Sorry for the late response, it always takes me some time to get my head around the CRMarchaeo issues;-).

1. Great that we can work together CRM-EH/CRMarchaeo compatibility

2. Question on “How Cuts, Fills and Interfaces actually relate”

We moved the type “Cut, Fill, ...” to the property “AP11.1 has type” which is attached to “AP11 has physical relation”, as displayed in Figure 5 and 7. So A3 [3] can have several “physical relations” of type “cuts” to other “A8 Stratigraphic Units” like illustrated in Figure 5.

Apart from that an A3 may “AP12 confine” several A2 Volume Units (We came up with “Volume” instead of “Deposit” based on your earlier comments, what is your opinion?).

Do you think this model can describe the situation or are we missing something? Maybe we have to try to model all the relations in the example and then see if something is missing.

3. “One event of removal process may produce many stratigraphic interfaces - I wouldn’t say so. It produces One stratigraphic interface, and Many physical interfaces. 1 stratigraphic relation = 1 event (of uncertain duration). I think that’s pretty fundamental to stratigraphic recording.”

I believe you are right here and we have to correct that.

4. Maybe we should try to model an example with Pauls latest model and with CRMarchaeo to see what’s missing or if there are places where we are uncertain if the semantics are correct.

Does anybody has a suggestion how we proceed organisationally with the review process of CRMarchaeo. Some people are doing bits and pieces, but we probably need a place where to put the latest version for review and comments and maybe somebody to organise the collaboration.

Posted by keith May on 12/11/2014

Hi Gerald,

I note and agree your final point “but we probably need a place where to put the latest version for review and comments and maybe somebody to organise the collaboration.”

I guess a place that shows the latest reviewed version is something ARIADNE partners will need too in due course?

1. Yes, and should be easier if we get a place to share the ‘latest/current’ version.

3. Thanks for confirmation

4. Sounds good but not sure how much time/resource is available in the near future (i.e. Paul and me), but again a shared review space might help with that too ?

2. With regard to choosing how to express the properties of Physical Relationships:

I believe what we need is a way to explicitly ‘semantically’ express that some Physical Relationships only ‘make sense’ between A2s (deposits/volumes) and A3s (cuts interfaces) while some only ‘make sense’ between A2s and A2s; and some only between A3 (cut interface) and both A2s – A3s.

So archaeologically:

A2 (deposit) Fills A3 (cut interface) – but cannot be a relationship with A2 i.e. A2 (deposit) cannot Fill an A2 (deposit)

A3 (cut interface) Is Filled by A2 (deposit) – but cannot be a relationship with A3 i.e. A3 (cut) cannot ‘is Filled by’ an A3 (cut)

A3 Cuts A2 or A3 – A3(cut) can ‘Cuts’ either A2 (deposit) or A3 (cut)

A2 or A3 Is Cut by A3 – cannot have a ‘is cut by’ relationship with A2 (deposit)

A2 (wall) Is bonded with A2 (wall) – cannot have a ‘Is bonded with’ relationship to A3

A2 (wall) Butts A2 (wall) – cannot have a ‘Butts’ relationship to A3

A2 (wall) Butted By A2 (wall) – cannot have a ‘Butted by’ relationship to A3

A2 Jointed with A2 (timber) – cannot have a ‘Jointed with’ relationship to A3

In an email exchange with Ceri we considered the ways to include the semantics of the above and he sent the following outline of how to express the relationships in (generalised) RDF.

Yes with sub-properties you can declare the type of thing permissible on both sides of the relationship (e.g.):

archaeo:AP11c_fills

rdfs:domain archaeo:A2_Deposit;

rdfs:range archaeo:A3_cut_interface .

You can’t do that with the ‘property of a property’ approach.

I think it does matter, otherwise may just as well use a totally generic model with zero semantic implications:

:Fred has_property (has_type=’name’) “Fred”;

has_property (has_type=’age’) “52”;

has_property (has_type=’gender’) “male”.

Etc.

A slightly more complete answer, although only to the points you raised, not necessarily an exhaustive model of all possible physical relationships

– no property number identifiers but you get the idea:

@prefix rdf: <http://www.w3.org/1999/02/22-rdf-syntax-ns#> .

@prefix rdfs: <http://www.w3.org/2000/01/rdf-schema#> .

@prefix owl: <http://www.w3.org/2002/07/owl#> .

@prefix archaeo: <http://www.ics.forth.gr/isl/CRMarchaeo/> .

archaeo:fills rdfs:label "fills"@en ;