Re: [Crm-sig] Modelling .1 properties

[martin]

Dear Simon,

Of course reification is natural. The argument of the Group against 
reification  for the .1 properties is the confusion of agency:


Reification is a statement about a statement. It implies an agent 
different from the statement it is about.
 A CIDOC CRM knowledge base must have a defauflt provenance "who says 
it". The .1 properties do not have a
different source of knowledge. Using the reification just because of a 
lack of RDFS of typed relations is, to my
opinion, a hack. It confuses ontology with syntax. The difference 
becomes important when we describe annotations to argue about
possibly different believes, i.e., when the default actor of the 
knowledge base makes statements about other's
believes or sources of knowledge. We came to this conclusion after 
studying the logic of shortcuts.


Would that make sense :-) ?

Best,

Martin

On 3/2/2016 11:47 μμ, Simon Spero wrote:


There is generally no problem with reification qua reification!
The CRM is committed to reification in its model of actions and 
events. See Davidson's "The Logical Form of Action Sentences" [1] for 
the paper that made this approach dominant. It's worth a read.


The collection "Essays on Action and Events" [2] collects other 
related papers; for  commentary see [3] -  mostly for Quine's essay on 
identity conditions for events to which Davidson is replying in the 
second edition of [2].  Identity is one of the  things that requires 
consideration when considering reification.


-

What is usually being objected to specifically is "RDF reification".
This facility consists of a class called rdf:Statement, together with 
the properties - subject, predicate, and object.


The mechanism that is provided is pretty awful.

Here is a reified statement.

_:x a rdf:Statement,
 rdf:subject :foo,
 rdf:predicate :property,
 rdf:object :bar .

This does *not* entail
:foo :property :bar.



[1] 
http://isites.harvard.edu/fs/docs/icb.topic638346.files/Davidson1967.pdf


[2] 
http://www.oxfordscholarship.com/view/10.1093/0199246270.001.0001/acprof-9780199246274


[3] 
https://books.google.com/books/about/Actions_and_Events.html?id=Tjl6QgAACAAJ&source=kp_cover


On Feb 3, 2016 2:02 PM, "Dan Matei" <mailto:d...@cimec.ro>> wrote:


Despite the fact that I'm always told that reification is not
recomanded, I decided to "piser contre le vent" :-)

Functionally, the reification is just natural. Proof: the
"invention" of .1 properties in CRM.

Aaa, if the formalisms we have do not handle it well, please
invent a suitable formalism, my dear friends.

Dan

--
Dan Matei
consultant (documentaristică, biblioteci digitale),
Fundația Gellu Naum,
[Institutul Național al Patrimoniului]


-Original Message-
From: martin mailto:mar...@ics.forth.gr>>
To: crm-sig@ics.forth.gr <mailto:crm-sig@ics.forth.gr>
Date: Wed, 3 Feb 2016 20:09:26 +0200
Subject: Re: [Crm-sig] Modelling .1 properties

> Dear Simon,
>
> Our messages crossed, your analysis is correct! We have discussed
> however, that reification or annotation is not recommended,
rather an
> introduction of a node (class) representing a triary property
rather than
> an individual entity. From the existence of an instance of
> PC14_carried_out_by we can automatically infer
> the instance of P14, as described in the formalization.
>
> All the best,
>
> martin
>
> On 3/2/2016 7:55 μμ, Simon Spero wrote:
> >
> > The first order formalization  given in the crm document are:
> >
> > P14(x,y,z) ⊃ [P14(x,y) ∧ E55(z)]
> >
> > Note that the predicate on the left hand side has three arguments,
> > which is more arguments than rdf is comfortable with.
> >
> > The "in the role of" property is modifying an instance of a
"carried
> > out by" property.
> > An activity can be carried out by several different agents,
each in a
> > different role, so the property cannot be attached directly to the
> > activity.
> >
> > There are several possible ways of representing this using
semantic
> > web tools.
> >
> > The first approach is to use RDF reification. I am not going
to say
> > anything more about this.
> >
> > If you are using OWL 2, you can add an annotation each
"carried out
> > by" property assertion.  This is not ideal, as annotations are not
> > really supposed to be part of the data in the model, and most
> > reasoners ignore them. They a

Re: [Crm-sig] Modelling .1 properties

[Chryssoula Bekiari]

Dear All
You may find the discussion about these properties and the final 
decision  in the issue 266.

http://83.212.168.219/CIDOC-CRM/Issue/reified-association-vs-sub-event

best
Chryssoula

On 3/2/2016 7:47 μμ, martin wrote:

Dear Allison,

The standard recommendation is to add a local vocabulary of 
subproperties of P14, if you have
a fixed vocabulary of roles. If not, CRM SIG is about to recommend the 
solution discussed in

http://www.cidoc-crm.org/docs/2015Feb02-32nd-meeting-minutes.pdf page 8:
introducing for each propertty having a .1 property a new class: 
"PC14_carried_out_by" etc.
that carries the .1 property and connects to domain and range of the 
base property.


We are stll lagging behind producing the corresponding RDFS :-[ .

My colleagues can send you a draft implementation.

Best,

Martin



On 3/2/2016 4:50 μμ, Allison Miller wrote:


Hello,

I wish to use the CIDOC CRM but need a formal model to do so and have 
a question concerning .1 properties.


eg. E7 Activity has: P14 carried out by (performed): E39 Actor and 
(P14.1 in the role of: E55 Type)


It is a property I need to use - but I can’t work out how to model it!

I thought I could use the Erlangen OWL implementation, but I can’t 
find these properties in it. (That’s not to claim they aren’t there, 
my knowledge of OWL is limited.)


I would welcome any guidance on P14.1, and other .1 properties, in 
the Erlangen implementation, or advice on including them in a 
definition compatible with Semantic Web technologies if anyone has 
done this.


Kind regards,

Alli

E-mail: allison.mil...@sysemia.co.uk

Web: www.sysemia.com



Sysemia Limited

The Innovation Centre, Bristol & Bath Science Park, Dirac Crescent, 
Emerson's Green, Bristol BS16 7FR


Registered in England and Wales. Company Number: 7555456

DISCLAIMER

Information contained in this e-mail is intended for the use of the 
addressee only, and is confidential and may also be privileged. If 
you receive this message in error, please advise us immediately. If 
you are not the intended recipient(s), please note that any form of 
distribution, copying or use of this communication or the information 
in it is strictly prohibited and may be unlawful. Attachments to this 
e-mail may contain software viruses which may damage your systems. 
Sysemia Ltd have taken reasonable steps to minimise this risk, but we 
advise that any attachments are virus checked before they are opened.




___
Crm-sig mailing list
Crm-sig@ics.forth.gr
http://lists.ics.forth.gr/mailman/listinfo/crm-sig



--

--
  Dr. Martin Doerr  |  Vox:+30(2810)391625|
  Research Director |  Fax:+30(2810)391638|
|  Email:mar...@ics.forth.gr  |
  |
Center for Cultural Informatics   |
Information Systems Laboratory|
 Institute of Computer Science|
Foundation for Research and Technology - Hellas (FORTH)   |
  |
N.Plastira 100, Vassilika Vouton, |
 GR70013 Heraklion,Crete,Greece   |
  |
  Web-site:http://www.ics.forth.gr/isl|
--




--
--
Chryssoula Bekiari
Research and Development Engineer

Center for Cultural Informatics / Information Systems Laboratory
Institute of Computer Science
Foundation for Research and Technology - Hellas (FORTH)

N. Plastira 100, Vassilika Vouton, GR-700 13 Heraklion, Crete, Greece
Phone: +30 2810 391631, Fax: +30 2810 391638, Skype: xrysmp
E-mail: beki...@ics.forth.gr

Web-site: http://www.ics.forth.gr/isl/people/people_individual.jsp?Person_ID=13
-

Re: [Crm-sig] Modelling .1 properties

[Simon Spero]

It is easy to specify a property value whose subject is a Property  - this
is just a simple case of OWL 2 punning.  However this is not the same as
applying a property to a triple, or to specify properties of properties.

Sometimes the problem is the inability of rdf to handle predicates with
more than two arguments. Thus a predicate that would requires three
arguments must be reified; in a different ontology language the three
argument predicate could be expressed directly.

In other cases, the predicate may not have a finite arity. This is the
problem that Davidson was addressing with reified events.

Some ontology languages allow for an entirely different kind of  properties
of properties. Languages like CycL, Common Logic, and KIF allow you to
define predicates that take other predicates as arguments. These systems
are not true 2nd order logics, as they restrict the range of predicates to
those explicitly mentioned, but they allow things like reflexivety or
transitivity of various kinds to be expressed in the language, rather than
requiring such concepts to be primitive.

This kind of expressivity can be  very useful when trying to express how
various properties of an item in a collection might be derived from
properties of the collection (or vice versa). This is especially useful
when combined with modal, probabilistic, or non-montonic reasoning (for
example, many properties of a FRBR item can be derived from its
manifestation ; however some items may differ from the prototype - pieces
may be missing, or the copy signed).

Simon

On Thu, Feb 4, 2016, 3:30 PM Dan Matei  wrote:

> Hi Martin,
>
> So, the PCxxx classes are a just a contortion of properties. I guess they
> are useful only because rdfs and owl do not accept properties as domains.
> There are other practical uses of them ?
>
> Since I do not care much about RDF, the tuples in my database are
> something like:
>
><.1 predicateQualifierId> 
> 
>
>
> If the  is a crm:E62_String, then I have also .
>
> If the  is a crm:E60_Number or a crm:E61_Time_Primitive, I also
> have  (e.g. =, <= etc. circa, ante, nonpost etc.) and
> .
>
> Thus every predicate can have at least one .1 property. But I also can say
> (simplified):
>
>
> <#1> <#EiffelTower> <#P43_has_dimension> <#height> <#circa> "300" <#m>
> <#me:2016-02-04>
>
> <#2> <#1> <#isDoubtful> "true" <#me:2016-02-04>
>
>
> When I will have to (when ?), I hope I will be able to do a decent json-ld
> serialization.
>
> It is (too) heretical ?
>
>
> Dan
>
>
>
> On 3 February 2016 at 22:37, martin  wrote:
>
>> Hi Dan,
>>
>> The "PCXXX" classes are exactly the suitable formalism you ask for.,
>> graph-topologically equivalent to the reification.
>> We must not confuse the syntactic pattern, which is the same for the
>> PCXXX solution, the annotation
>> and the reification, with the intended meaning ;-) .
>>
>> Cheers,
>>
>> Martin
>>
>>
>> On 3/2/2016 8:58 μμ, Dan Matei wrote:
>>
>>> Despite the fact that I'm always told that reification is not
>>> recomanded, I decided to "piser contre le vent" :-)
>>>
>>> Functionally, the reification is just natural. Proof: the "invention" of
>>> .1 properties in CRM.
>>>
>>> Aaa, if the formalisms we have do not handle it well, please invent a
>>> suitable formalism, my dear friends.
>>>
>>> Dan
>>>
>>> --
>>> Dan Matei
>>> consultant (documentaristică, biblioteci digitale),
>>> Fundația Gellu Naum,
>>> [Institutul Național al Patrimoniului]
>>>
>>>
>>> -Original Message-
>>> From: martin 
>>> To: crm-sig@ics.forth.gr
>>> Date: Wed, 3 Feb 2016 20:09:26 +0200
>>> Subject: Re: [Crm-sig] Modelling .1 properties
>>>
>>> Dear Simon,
>>>>
>>>> Our messages crossed, your analysis is correct! We have discussed
>>>> however, that reification or annotation is not recommended, rather an
>>>> introduction of a node (class) representing a triary property rather
>>>> than
>>>> an individual entity. From the existence of an instance of
>>>> PC14_carried_out_by we can automatically infer
>>>> the instance of P14, as described in the formalization.
>>>>
>>>> All the best,
>>>>
>>>> martin
>>>>
>>>> On 3/2/2016 7:55 μμ, Simon Spero wrote:
>>>>
>>>>> The first or

Re: [Crm-sig] Modelling .1 properties

[Dan Matei]

Hi Simon,


On 3 February 2016 at 23:47, Simon Spero  wrote:

> The mechanism that is provided is pretty awful.
>
Agree :-)

> Here is a reified statement.
>
> _:x a rdf:Statement,
>  rdf:subject :foo,
>  rdf:predicate :property,
>  rdf:object :bar .
>
> This does *not* entail
> :foo :property :bar.
>
Yes, but if I understand that this entails :foo :property :bar, I can also
teach my software to understand also :-)

Dan

> 
>
> [1]
> http://isites.harvard.edu/fs/docs/icb.topic638346.files/Davidson1967.pdf
>
> [2]
> http://www.oxfordscholarship.com/view/10.1093/0199246270.001.0001/acprof-9780199246274
>
> [3]
> https://books.google.com/books/about/Actions_and_Events.html?id=Tjl6QgAACAAJ&source=kp_cover
> On Feb 3, 2016 2:02 PM, "Dan Matei"  wrote:
>
>> Despite the fact that I'm always told that reification is not recomanded,
>> I decided to "piser contre le vent" :-)
>>
>> Functionally, the reification is just natural. Proof: the "invention" of
>> .1 properties in CRM.
>>
>> Aaa, if the formalisms we have do not handle it well, please invent a
>> suitable formalism, my dear friends.
>>
>> Dan
>>
>> --
>> Dan Matei
>> consultant (documentaristică, biblioteci digitale),
>> Fundația Gellu Naum,
>> [Institutul Național al Patrimoniului]
>>
>>
>> -Original Message-
>> From: martin 
>> To: crm-sig@ics.forth.gr
>> Date: Wed, 3 Feb 2016 20:09:26 +0200
>> Subject: Re: [Crm-sig] Modelling .1 properties
>>
>> > Dear Simon,
>> >
>> > Our messages crossed, your analysis is correct! We have discussed
>> > however, that reification or annotation is not recommended, rather an
>> > introduction of a node (class) representing a triary property rather
>> than
>> > an individual entity. From the existence of an instance of
>> > PC14_carried_out_by we can automatically infer
>> > the instance of P14, as described in the formalization.
>> >
>> > All the best,
>> >
>> > martin
>> >
>> > On 3/2/2016 7:55 μμ, Simon Spero wrote:
>> > >
>> > > The first order formalization  given in the crm document are:
>> > >
>> > > P14(x,y,z) ⊃ [P14(x,y) ∧ E55(z)]
>> > >
>> > > Note that the predicate on the left hand side has three arguments,
>> > > which is more arguments than rdf is comfortable with.
>> > >
>> > > The "in the role of" property is modifying an instance of a "carried
>> > > out by" property.
>> > > An activity can be carried out by several different agents, each in a
>> > > different role, so the property cannot be attached directly to the
>> > > activity.
>> > >
>> > > There are several possible ways of representing this using semantic
>> > > web tools.
>> > >
>> > > The first approach is to use RDF reification. I am not going to say
>> > > anything more about this.
>> > >
>> > > If you are using OWL 2, you can add an annotation each "carried out
>> > > by" property assertion.  This is not ideal, as annotations are not
>> > > really supposed to be part of the data in the model, and most
>> > > reasoners ignore them. They are also not easy to work with in RDF.
>> > >
>> > > A third approach is to define your own class for reification,
>> > > representing an instance of a "carrying out" ; this class would have
>> > > properties relating the activity, the agent, and the role.
>> > >
>> > > The best approach may be to define a sub property of P14 for each type
>> > > of carrying out in a role which is relevant to your model.
>> > > You can specify the role associated with all uses of this property
>> > > using a property whose subject is the subproperty.
>> > >
>> > > Simon
>> > >
>> > > On Feb 3, 2016 10:01 AM, "Allison Miller"
>> > > mailto:allison.mil...@sysemia.co.uk>>
>> > > wrote:
>> > >
>> > > Hello,
>> > >
>> > > I wish to use the CIDOC CRM but need a formal model to do so and
>> > > have a question concerning .1 properties.
>> > >
>> > > eg. E7 Activity has: P14 carried out by (performed): E39 Actor and
>> &

Re: [Crm-sig] Modelling .1 properties

[Dan Matei]

Hi Martin,

So, the PCxxx classes are a just a contortion of properties. I guess they
are useful only because rdfs and owl do not accept properties as domains.
There are other practical uses of them ?

Since I do not care much about RDF, the tuples in my database are something
like:

   <.1 predicateQualifierId> 



If the  is a crm:E62_String, then I have also .

If the  is a crm:E60_Number or a crm:E61_Time_Primitive, I also
have  (e.g. =, <= etc. circa, ante, nonpost etc.) and
.

Thus every predicate can have at least one .1 property. But I also can say
(simplified):


<#1> <#EiffelTower> <#P43_has_dimension> <#height> <#circa> "300" <#m>
<#me:2016-02-04>

<#2> <#1> <#isDoubtful> "true" <#me:2016-02-04>


When I will have to (when ?), I hope I will be able to do a decent json-ld
serialization.

It is (too) heretical ?


Dan



On 3 February 2016 at 22:37, martin  wrote:

> Hi Dan,
>
> The "PCXXX" classes are exactly the suitable formalism you ask for.,
> graph-topologically equivalent to the reification.
> We must not confuse the syntactic pattern, which is the same for the PCXXX
> solution, the annotation
> and the reification, with the intended meaning ;-) .
>
> Cheers,
>
> Martin
>
>
> On 3/2/2016 8:58 μμ, Dan Matei wrote:
>
>> Despite the fact that I'm always told that reification is not recomanded,
>> I decided to "piser contre le vent" :-)
>>
>> Functionally, the reification is just natural. Proof: the "invention" of
>> .1 properties in CRM.
>>
>> Aaa, if the formalisms we have do not handle it well, please invent a
>> suitable formalism, my dear friends.
>>
>> Dan
>>
>> --
>> Dan Matei
>> consultant (documentaristică, biblioteci digitale),
>> Fundația Gellu Naum,
>> [Institutul Național al Patrimoniului]
>>
>>
>> -Original Message-
>> From: martin 
>> To: crm-sig@ics.forth.gr
>> Date: Wed, 3 Feb 2016 20:09:26 +0200
>> Subject: Re: [Crm-sig] Modelling .1 properties
>>
>> Dear Simon,
>>>
>>> Our messages crossed, your analysis is correct! We have discussed
>>> however, that reification or annotation is not recommended, rather an
>>> introduction of a node (class) representing a triary property rather than
>>> an individual entity. From the existence of an instance of
>>> PC14_carried_out_by we can automatically infer
>>> the instance of P14, as described in the formalization.
>>>
>>> All the best,
>>>
>>> martin
>>>
>>> On 3/2/2016 7:55 μμ, Simon Spero wrote:
>>>
>>>> The first order formalization  given in the crm document are:
>>>>
>>>> P14(x,y,z) ⊃ [P14(x,y) ∧ E55(z)]
>>>>
>>>> Note that the predicate on the left hand side has three arguments,
>>>> which is more arguments than rdf is comfortable with.
>>>>
>>>> The "in the role of" property is modifying an instance of a "carried
>>>> out by" property.
>>>> An activity can be carried out by several different agents, each in a
>>>> different role, so the property cannot be attached directly to the
>>>> activity.
>>>>
>>>> There are several possible ways of representing this using semantic
>>>> web tools.
>>>>
>>>> The first approach is to use RDF reification. I am not going to say
>>>> anything more about this.
>>>>
>>>> If you are using OWL 2, you can add an annotation each "carried out
>>>> by" property assertion.  This is not ideal, as annotations are not
>>>> really supposed to be part of the data in the model, and most
>>>> reasoners ignore them. They are also not easy to work with in RDF.
>>>>
>>>> A third approach is to define your own class for reification,
>>>> representing an instance of a "carrying out" ; this class would have
>>>> properties relating the activity, the agent, and the role.
>>>>
>>>> The best approach may be to define a sub property of P14 for each type
>>>> of carrying out in a role which is relevant to your model.
>>>> You can specify the role associated with all uses of this property
>>>> using a property whose subject is the subproperty.
>>>>
>>>> Simon
>>>>
>>>> On Feb 3, 2016 10:01 AM, "Allison Miller"
>>>> m

Re: [Crm-sig] Modelling .1 properties

[Maria Theodoridou]

Dear all,

Attached you can find the rdfs encoding for the .1 properties as Martin 
described.
Also in the file http://www.ics.forth.gr/isl/CRMext/Roles.pdf you can 
see 3 possible solutions and relevant examples.


The rdfs has not been tested extensively, so if you notice any errors 
please let me know.


Best,
Maria


On 3/2/2016 11:47 μμ, Simon Spero wrote:


There is generally no problem with reification qua reification!
The CRM is committed to reification in its model of actions and 
events. See Davidson's "The Logical Form of Action Sentences" [1] for 
the paper that made this approach dominant. It's worth a read.


The collection "Essays on Action and Events" [2] collects other 
related papers; for  commentary see [3] -  mostly for Quine's essay on 
identity conditions for events to which Davidson is replying in the 
second edition of [2].  Identity is one of the  things that requires 
consideration when considering reification.


-

What is usually being objected to specifically is "RDF reification".
This facility consists of a class called rdf:Statement, together with 
the properties - subject, predicate, and object.


The mechanism that is provided is pretty awful.

Here is a reified statement.

_:x a rdf:Statement,
 rdf:subject :foo,
 rdf:predicate :property,
 rdf:object :bar .

This does *not* entail
:foo :property :bar.



[1] 
http://isites.harvard.edu/fs/docs/icb.topic638346.files/Davidson1967.pdf


[2] 
http://www.oxfordscholarship.com/view/10.1093/0199246270.001.0001/acprof-9780199246274


[3] 
https://books.google.com/books/about/Actions_and_Events.html?id=Tjl6QgAACAAJ&source=kp_cover


On Feb 3, 2016 2:02 PM, "Dan Matei" <mailto:d...@cimec.ro>> wrote:


Despite the fact that I'm always told that reification is not
recomanded, I decided to "piser contre le vent" :-)

Functionally, the reification is just natural. Proof: the
"invention" of .1 properties in CRM.

Aaa, if the formalisms we have do not handle it well, please
invent a suitable formalism, my dear friends.

Dan

--
Dan Matei
consultant (documentaristică, biblioteci digitale),
Fundația Gellu Naum,
[Institutul Național al Patrimoniului]


-Original Message-
From: martin mailto:mar...@ics.forth.gr>>
To: crm-sig@ics.forth.gr <mailto:crm-sig@ics.forth.gr>
Date: Wed, 3 Feb 2016 20:09:26 +0200
Subject: Re: [Crm-sig] Modelling .1 properties

> Dear Simon,
>
> Our messages crossed, your analysis is correct! We have discussed
> however, that reification or annotation is not recommended,
rather an
> introduction of a node (class) representing a triary property
rather than
> an individual entity. From the existence of an instance of
> PC14_carried_out_by we can automatically infer
> the instance of P14, as described in the formalization.
>
> All the best,
>
> martin
>
> On 3/2/2016 7:55 μμ, Simon Spero wrote:
> >
> > The first order formalization  given in the crm document are:
> >
> > P14(x,y,z) ⊃ [P14(x,y) ∧ E55(z)]
> >
> > Note that the predicate on the left hand side has three arguments,
> > which is more arguments than rdf is comfortable with.
> >
> > The "in the role of" property is modifying an instance of a
"carried
> > out by" property.
> > An activity can be carried out by several different agents,
each in a
> > different role, so the property cannot be attached directly to the
> > activity.
> >
> > There are several possible ways of representing this using
semantic
> > web tools.
> >
> > The first approach is to use RDF reification. I am not going
to say
> > anything more about this.
> >
> > If you are using OWL 2, you can add an annotation each
"carried out
> > by" property assertion.  This is not ideal, as annotations are not
> > really supposed to be part of the data in the model, and most
> > reasoners ignore them. They are also not easy to work with in RDF.
> >
> > A third approach is to define your own class for reification,
> > representing an instance of a "carrying out" ; this class
would have
> > properties relating the activity, the agent, and the role.
> >
> > The best approach may be to define a sub property of P14 for
each type
> > of carrying out in a role which is relevant to your model.
> > You can specify the role associated with all uses of this property
   

Re: [Crm-sig] Modelling .1 properties

[Simon Spero]

There is generally no problem with reification qua reification!
The CRM is committed to reification in its model of actions and events. See
Davidson's "The Logical Form of Action Sentences" [1] for the paper that
made this approach dominant. It's worth a read.

The collection "Essays on Action and Events" [2] collects other related
papers; for  commentary see [3] -  mostly for Quine's essay on identity
conditions for events to which Davidson is replying in the second edition
of [2].  Identity is one of the  things that requires consideration when
considering reification.

-

What is usually being objected to specifically is "RDF reification".
This facility consists of a class called rdf:Statement, together with the
properties - subject, predicate, and object.

The mechanism that is provided is pretty awful.

Here is a reified statement.

_:x a rdf:Statement,
 rdf:subject :foo,
 rdf:predicate :property,
 rdf:object :bar .

This does *not* entail
:foo :property :bar.



[1] http://isites.harvard.edu/fs/docs/icb.topic638346.files/Davidson1967.pdf

[2]
http://www.oxfordscholarship.com/view/10.1093/0199246270.001.0001/acprof-9780199246274

[3]
https://books.google.com/books/about/Actions_and_Events.html?id=Tjl6QgAACAAJ&source=kp_cover
On Feb 3, 2016 2:02 PM, "Dan Matei"  wrote:

> Despite the fact that I'm always told that reification is not recomanded,
> I decided to "piser contre le vent" :-)
>
> Functionally, the reification is just natural. Proof: the "invention" of
> .1 properties in CRM.
>
> Aaa, if the formalisms we have do not handle it well, please invent a
> suitable formalism, my dear friends.
>
> Dan
>
> --
> Dan Matei
> consultant (documentaristică, biblioteci digitale),
> Fundația Gellu Naum,
> [Institutul Național al Patrimoniului]
>
>
> -----Original Message-
> From: martin 
> To: crm-sig@ics.forth.gr
> Date: Wed, 3 Feb 2016 20:09:26 +0200
> Subject: Re: [Crm-sig] Modelling .1 properties
>
> > Dear Simon,
> >
> > Our messages crossed, your analysis is correct! We have discussed
> > however, that reification or annotation is not recommended, rather an
> > introduction of a node (class) representing a triary property rather than
> > an individual entity. From the existence of an instance of
> > PC14_carried_out_by we can automatically infer
> > the instance of P14, as described in the formalization.
> >
> > All the best,
> >
> > martin
> >
> > On 3/2/2016 7:55 μμ, Simon Spero wrote:
> > >
> > > The first order formalization  given in the crm document are:
> > >
> > > P14(x,y,z) ⊃ [P14(x,y) ∧ E55(z)]
> > >
> > > Note that the predicate on the left hand side has three arguments,
> > > which is more arguments than rdf is comfortable with.
> > >
> > > The "in the role of" property is modifying an instance of a "carried
> > > out by" property.
> > > An activity can be carried out by several different agents, each in a
> > > different role, so the property cannot be attached directly to the
> > > activity.
> > >
> > > There are several possible ways of representing this using semantic
> > > web tools.
> > >
> > > The first approach is to use RDF reification. I am not going to say
> > > anything more about this.
> > >
> > > If you are using OWL 2, you can add an annotation each "carried out
> > > by" property assertion.  This is not ideal, as annotations are not
> > > really supposed to be part of the data in the model, and most
> > > reasoners ignore them. They are also not easy to work with in RDF.
> > >
> > > A third approach is to define your own class for reification,
> > > representing an instance of a "carrying out" ; this class would have
> > > properties relating the activity, the agent, and the role.
> > >
> > > The best approach may be to define a sub property of P14 for each type
> > > of carrying out in a role which is relevant to your model.
> > > You can specify the role associated with all uses of this property
> > > using a property whose subject is the subproperty.
> > >
> > > Simon
> > >
> > > On Feb 3, 2016 10:01 AM, "Allison Miller"
> > > mailto:allison.mil...@sysemia.co.uk>>
> > > wrote:
> > >
> > > Hello,
> > >
> > > I wish to use the CIDOC CRM but need a formal model to do so and
> > > h

Re: [Crm-sig] Modelling .1 properties

[martin]

Hi Dan,

The "PCXXX" classes are exactly the suitable formalism you ask for., 
graph-topologically equivalent to the reification.
We must not confuse the syntactic pattern, which is the same for the 
PCXXX solution, the annotation

and the reification, with the intended meaning ;-) .

Cheers,

Martin

On 3/2/2016 8:58 μμ, Dan Matei wrote:

Despite the fact that I'm always told that reification is not recomanded, I decided to 
"piser contre le vent" :-)

Functionally, the reification is just natural. Proof: the "invention" of .1 
properties in CRM.

Aaa, if the formalisms we have do not handle it well, please invent a suitable 
formalism, my dear friends.

Dan

--
Dan Matei
consultant (documentaristică, biblioteci digitale),
Fundația Gellu Naum,
[Institutul Național al Patrimoniului]


-Original Message-
From: martin 
To: crm-sig@ics.forth.gr
Date: Wed, 3 Feb 2016 20:09:26 +0200
Subject: Re: [Crm-sig] Modelling .1 properties


Dear Simon,

Our messages crossed, your analysis is correct! We have discussed
however, that reification or annotation is not recommended, rather an
introduction of a node (class) representing a triary property rather than
an individual entity. From the existence of an instance of
PC14_carried_out_by we can automatically infer
the instance of P14, as described in the formalization.

All the best,

martin

On 3/2/2016 7:55 μμ, Simon Spero wrote:

The first order formalization  given in the crm document are:

P14(x,y,z) ⊃ [P14(x,y) ∧ E55(z)]

Note that the predicate on the left hand side has three arguments,
which is more arguments than rdf is comfortable with.

The "in the role of" property is modifying an instance of a "carried
out by" property.
An activity can be carried out by several different agents, each in a
different role, so the property cannot be attached directly to the
activity.

There are several possible ways of representing this using semantic
web tools.

The first approach is to use RDF reification. I am not going to say
anything more about this.

If you are using OWL 2, you can add an annotation each "carried out
by" property assertion.  This is not ideal, as annotations are not
really supposed to be part of the data in the model, and most
reasoners ignore them. They are also not easy to work with in RDF.

A third approach is to define your own class for reification,
representing an instance of a "carrying out" ; this class would have
properties relating the activity, the agent, and the role.

The best approach may be to define a sub property of P14 for each type
of carrying out in a role which is relevant to your model.
You can specify the role associated with all uses of this property
using a property whose subject is the subproperty.

Simon

On Feb 3, 2016 10:01 AM, "Allison Miller"
mailto:allison.mil...@sysemia.co.uk>>
wrote:

 Hello,

 I wish to use the CIDOC CRM but need a formal model to do so and
 have a question concerning .1 properties.

 eg. E7 Activity has: P14 carried out by (performed): E39 Actor and
 (P14.1 in the role of: E55 Type)

 It is a property I need to use - but I can’t work out how to model it!

 I thought I could use the Erlangen OWL implementation, but I can’t
 find these properties in it. (That’s not to claim they aren’t
 there, my knowledge of OWL is limited.)

 I would welcome any guidance on P14.1, and other .1 properties, in
 the Erlangen implementation, or advice on including them in a
 definition compatible with Semantic Web technologies if anyone has
 done this.

 Kind regards,

 Alli

 E-mail: allison.mil...@sysemia.co.uk
 <mailto:allison.mil...@sysemia.co.uk>

 Web: www.sysemia.com <http://www.sysemia.com/>



 Sysemia Limited

 The Innovation Centre, Bristol & Bath Science Park, Dirac
 Crescent, Emerson's Green, Bristol BS16 7FR

 Registered in England and Wales. Company Number: 7555456

 DISCLAIMER

 Information contained in this e-mail is intended for the use of
 the addressee only, and is confidential and may also be
 privileged. If you receive this message in error, please advise us
 immediately. If you are not the intended recipient(s), please note
 that any form of distribution, copying or use of this
 communication or the information in it is strictly prohibited and
 may be unlawful. Attachments to this e-mail may contain software
 viruses which may damage your systems. Sysemia Ltd have taken
 reasonable steps to minimise this risk, but we advise that any
 attachments are virus checked before they are opened.


 ___
 Crm-sig mailing list
 Crm-sig@ics.forth.gr <mailto:Crm-sig@ics.forth.gr>
 http://lists.ics.forth.gr/mailman/listinfo/crm-sig



_

Re: [Crm-sig] Modelling .1 properties

[Dan Matei]

Despite the fact that I'm always told that reification is not recomanded, I 
decided to "piser contre le vent" :-)

Functionally, the reification is just natural. Proof: the "invention" of .1 
properties in CRM.

Aaa, if the formalisms we have do not handle it well, please invent a suitable 
formalism, my dear friends.

Dan

--
Dan Matei
consultant (documentaristică, biblioteci digitale),
Fundația Gellu Naum,
[Institutul Național al Patrimoniului]


-Original Message-
From: martin 
To: crm-sig@ics.forth.gr
List-Post: crm-sig@ics.forth.gr
Date: Wed, 3 Feb 2016 20:09:26 +0200
Subject: Re: [Crm-sig] Modelling .1 properties

> Dear Simon,
> 
> Our messages crossed, your analysis is correct! We have discussed 
> however, that reification or annotation is not recommended, rather an 
> introduction of a node (class) representing a triary property rather than
> an individual entity. From the existence of an instance of 
> PC14_carried_out_by we can automatically infer
> the instance of P14, as described in the formalization.
> 
> All the best,
> 
> martin
> 
> On 3/2/2016 7:55 μμ, Simon Spero wrote:
> >
> > The first order formalization  given in the crm document are:
> >
> > P14(x,y,z) ⊃ [P14(x,y) ∧ E55(z)]
> >
> > Note that the predicate on the left hand side has three arguments, 
> > which is more arguments than rdf is comfortable with.
> >
> > The "in the role of" property is modifying an instance of a "carried 
> > out by" property.
> > An activity can be carried out by several different agents, each in a 
> > different role, so the property cannot be attached directly to the 
> > activity.
> >
> > There are several possible ways of representing this using semantic 
> > web tools.
> >
> > The first approach is to use RDF reification. I am not going to say 
> > anything more about this.
> >
> > If you are using OWL 2, you can add an annotation each "carried out 
> > by" property assertion.  This is not ideal, as annotations are not 
> > really supposed to be part of the data in the model, and most 
> > reasoners ignore them. They are also not easy to work with in RDF.
> >
> > A third approach is to define your own class for reification, 
> > representing an instance of a "carrying out" ; this class would have 
> > properties relating the activity, the agent, and the role.
> >
> > The best approach may be to define a sub property of P14 for each type 
> > of carrying out in a role which is relevant to your model.
> > You can specify the role associated with all uses of this property 
> > using a property whose subject is the subproperty.
> >
> > Simon
> >
> > On Feb 3, 2016 10:01 AM, "Allison Miller" 
> > mailto:allison.mil...@sysemia.co.uk>> 
> > wrote:
> >
> > Hello,
> >
> > I wish to use the CIDOC CRM but need a formal model to do so and
> > have a question concerning .1 properties.
> >
> > eg. E7 Activity has: P14 carried out by (performed): E39 Actor and
> > (P14.1 in the role of: E55 Type)
> >
> > It is a property I need to use - but I can’t work out how to model it!
> >
> > I thought I could use the Erlangen OWL implementation, but I can’t
> > find these properties in it. (That’s not to claim they aren’t
> > there, my knowledge of OWL is limited.)
> >
> > I would welcome any guidance on P14.1, and other .1 properties, in
> > the Erlangen implementation, or advice on including them in a
> > definition compatible with Semantic Web technologies if anyone has
> > done this.
> >
> > Kind regards,
> >
> > Alli
> >
> > E-mail: allison.mil...@sysemia.co.uk
> > <mailto:allison.mil...@sysemia.co.uk>
> >
> > Web: www.sysemia.com <http://www.sysemia.com/>
> >
> >
> >
> > Sysemia Limited
> >
> > The Innovation Centre, Bristol & Bath Science Park, Dirac
> > Crescent, Emerson's Green, Bristol BS16 7FR
> >
> > Registered in England and Wales. Company Number: 7555456
> >
> > DISCLAIMER
> >
> > Information contained in this e-mail is intended for the use of
> > the addressee only, and is confidential and may also be
> > privileged. If you receive this message in error, please advise us
> > immediately. If you are not the intended recipient(s), please note
> > that any form of distribution, copying or 

Re: [Crm-sig] Modelling .1 properties

[martin]

Dear Simon,

Our messages crossed, your analysis is correct! We have discussed 
however, that reification or annotation is not recommended, rather an 
introduction of a node (class) representing a triary property rather than
an individual entity. From the existence of an instance of 
PC14_carried_out_by we can automatically infer

the instance of P14, as described in the formalization.

All the best,

martin

On 3/2/2016 7:55 μμ, Simon Spero wrote:


The first order formalization  given in the crm document are:

P14(x,y,z) ⊃ [P14(x,y) ∧ E55(z)]

Note that the predicate on the left hand side has three arguments, 
which is more arguments than rdf is comfortable with.


The "in the role of" property is modifying an instance of a "carried 
out by" property.
An activity can be carried out by several different agents, each in a 
different role, so the property cannot be attached directly to the 
activity.


There are several possible ways of representing this using semantic 
web tools.


The first approach is to use RDF reification. I am not going to say 
anything more about this.


If you are using OWL 2, you can add an annotation each "carried out 
by" property assertion.  This is not ideal, as annotations are not 
really supposed to be part of the data in the model, and most 
reasoners ignore them. They are also not easy to work with in RDF.


A third approach is to define your own class for reification, 
representing an instance of a "carrying out" ; this class would have 
properties relating the activity, the agent, and the role.


The best approach may be to define a sub property of P14 for each type 
of carrying out in a role which is relevant to your model.
You can specify the role associated with all uses of this property 
using a property whose subject is the subproperty.


Simon

On Feb 3, 2016 10:01 AM, "Allison Miller" 
mailto:allison.mil...@sysemia.co.uk>> 
wrote:


Hello,

I wish to use the CIDOC CRM but need a formal model to do so and
have a question concerning .1 properties.

eg. E7 Activity has: P14 carried out by (performed): E39 Actor and
(P14.1 in the role of: E55 Type)

It is a property I need to use - but I can’t work out how to model it!

I thought I could use the Erlangen OWL implementation, but I can’t
find these properties in it. (That’s not to claim they aren’t
there, my knowledge of OWL is limited.)

I would welcome any guidance on P14.1, and other .1 properties, in
the Erlangen implementation, or advice on including them in a
definition compatible with Semantic Web technologies if anyone has
done this.

Kind regards,

Alli

E-mail: allison.mil...@sysemia.co.uk


Web: www.sysemia.com 



Sysemia Limited

The Innovation Centre, Bristol & Bath Science Park, Dirac
Crescent, Emerson's Green, Bristol BS16 7FR

Registered in England and Wales. Company Number: 7555456

DISCLAIMER

Information contained in this e-mail is intended for the use of
the addressee only, and is confidential and may also be
privileged. If you receive this message in error, please advise us
immediately. If you are not the intended recipient(s), please note
that any form of distribution, copying or use of this
communication or the information in it is strictly prohibited and
may be unlawful. Attachments to this e-mail may contain software
viruses which may damage your systems. Sysemia Ltd have taken
reasonable steps to minimise this risk, but we advise that any
attachments are virus checked before they are opened.


___
Crm-sig mailing list
Crm-sig@ics.forth.gr 
http://lists.ics.forth.gr/mailman/listinfo/crm-sig



___
Crm-sig mailing list
Crm-sig@ics.forth.gr
http://lists.ics.forth.gr/mailman/listinfo/crm-sig



--

--
 Dr. Martin Doerr  |  Vox:+30(2810)391625|
 Research Director |  Fax:+30(2810)391638|
   |  Email: mar...@ics.forth.gr |
 |
   Center for Cultural Informatics   |
   Information Systems Laboratory|
Institute of Computer Science|
   Foundation for Research and Technology - Hellas (FORTH)   |
 |
   N.Plastira 100, Vassilika Vouton, |
GR70013 Heraklion,Crete,Greece   |
 |
 Web-site: http://www.ics.forth.gr/isl   |
--

Re: [Crm-sig] Modelling .1 properties

[Simon Spero]

The first order formalization  given in the crm document are:

P14(x,y,z) ⊃ [P14(x,y) ∧ E55(z)]

Note that the predicate on the left hand side has three arguments, which is
more arguments than rdf is comfortable with.

The "in the role of" property is modifying an instance of a "carried out
by" property.
An activity can be carried out by several different agents, each in a
different role, so the property cannot be attached directly to the
activity.

There are several possible ways of representing this using semantic web
tools.

The first approach is to use RDF reification. I am not going to say
anything more about this.

If you are using OWL 2, you can add an annotation each "carried out by"
property assertion.  This is not ideal, as annotations are not really
supposed to be part of the data in the model, and most reasoners ignore
them. They are also not easy to work with in RDF.

A third approach is to define your own class for reification, representing
an instance of a "carrying out" ; this class would have properties relating
the activity, the agent, and the role.

The best approach may be to define a sub property of P14 for each type of
carrying out in a role which is relevant to your model.
You can specify the role associated with all uses of this property using a
property whose subject is the subproperty.

Simon
On Feb 3, 2016 10:01 AM, "Allison Miller" 
wrote:

> Hello,
>
>
>
> I wish to use the CIDOC CRM but need a formal model to do so and have a
> question concerning .1 properties.
>
> eg. E7 Activity has: P14 carried out by (performed): E39 Actor and (P14.1
> in the role of: E55 Type)
>
>
>
> It is a property I need to use - but I can’t work out how to model it!
>
> I thought I could use the Erlangen OWL implementation, but I can’t find
> these properties in it. (That’s not to claim they aren’t there, my
> knowledge of OWL is limited.)
>
> I would welcome any guidance on P14.1, and other .1 properties, in the
> Erlangen implementation, or advice on including them in a definition
> compatible with Semantic Web technologies if anyone has done this.
>
>
>
> Kind regards,
>
> Alli
>
>
>
> E-mail: allison.mil...@sysemia.co.uk
>
> Web: www.sysemia.com
>
>
>
>
> Sysemia Limited
>
> The Innovation Centre, Bristol & Bath Science Park, Dirac Crescent,
> Emerson's Green, Bristol BS16 7FR
>
> Registered in England and Wales. Company Number: 7555456
>
>
>
> DISCLAIMER
>
> Information contained in this e-mail is intended for the use of the
> addressee only, and is confidential and may also be privileged. If you
> receive this message in error, please advise us immediately. If you are not
> the intended recipient(s), please note that any form of distribution,
> copying or use of this communication or the information in it is strictly
> prohibited and may be unlawful. Attachments to this e-mail may contain
> software viruses which may damage your systems. Sysemia Ltd have taken
> reasonable steps to minimise this risk, but we advise that any attachments
> are virus checked before they are opened.
>
> ___
> Crm-sig mailing list
> Crm-sig@ics.forth.gr
> http://lists.ics.forth.gr/mailman/listinfo/crm-sig
>
>

Re: [Crm-sig] Modelling .1 properties

[martin]

Dear Allison,

The standard recommendation is to add a local vocabulary of 
subproperties of P14, if you have
a fixed vocabulary of roles. If not, CRM SIG is about to recommend the 
solution discussed in

http://www.cidoc-crm.org/docs/2015Feb02-32nd-meeting-minutes.pdf page 8:
introducing for each propertty having a .1 property a new class: 
"PC14_carried_out_by" etc.
that carries the .1 property and connects to domain and range of the 
base property.


We are stll lagging behind producing the corresponding RDFS :-[ .

My colleagues can send you a draft implementation.

Best,

Martin



On 3/2/2016 4:50 μμ, Allison Miller wrote:


Hello,

I wish to use the CIDOC CRM but need a formal model to do so and have 
a question concerning .1 properties.


eg. E7 Activity has: P14 carried out by (performed): E39 Actor and 
(P14.1 in the role of: E55 Type)


It is a property I need to use - but I can’t work out how to model it!

I thought I could use the Erlangen OWL implementation, but I can’t 
find these properties in it. (That’s not to claim they aren’t there, 
my knowledge of OWL is limited.)


I would welcome any guidance on P14.1, and other .1 properties, in the 
Erlangen implementation, or advice on including them in a definition 
compatible with Semantic Web technologies if anyone has done this.


Kind regards,

Alli

E-mail: allison.mil...@sysemia.co.uk 

Web: www.sysemia.com 



Sysemia Limited

The Innovation Centre, Bristol & Bath Science Park, Dirac Crescent, 
Emerson's Green, Bristol BS16 7FR


Registered in England and Wales. Company Number: 7555456

DISCLAIMER

Information contained in this e-mail is intended for the use of the 
addressee only, and is confidential and may also be privileged. If you 
receive this message in error, please advise us immediately. If you 
are not the intended recipient(s), please note that any form of 
distribution, copying or use of this communication or the information 
in it is strictly prohibited and may be unlawful. Attachments to this 
e-mail may contain software viruses which may damage your systems. 
Sysemia Ltd have taken reasonable steps to minimise this risk, but we 
advise that any attachments are virus checked before they are opened.




___
Crm-sig mailing list
Crm-sig@ics.forth.gr
http://lists.ics.forth.gr/mailman/listinfo/crm-sig



--

--
 Dr. Martin Doerr  |  Vox:+30(2810)391625|
 Research Director |  Fax:+30(2810)391638|
   |  Email: mar...@ics.forth.gr |
 |
   Center for Cultural Informatics   |
   Information Systems Laboratory|
Institute of Computer Science|
   Foundation for Research and Technology - Hellas (FORTH)   |
 |
   N.Plastira 100, Vassilika Vouton, |
GR70013 Heraklion,Crete,Greece   |
 |
 Web-site: http://www.ics.forth.gr/isl   |
--

[Crm-sig] Modelling .1 properties

[Allison Miller]

Hello,



I wish to use the CIDOC CRM but need a formal model to do so and have a
question concerning .1 properties.

eg. E7 Activity has: P14 carried out by (performed): E39 Actor and (P14.1
in the role of: E55 Type)



It is a property I need to use - but I can’t work out how to model it!

I thought I could use the Erlangen OWL implementation, but I can’t find
these properties in it. (That’s not to claim they aren’t there, my
knowledge of OWL is limited.)

I would welcome any guidance on P14.1, and other .1 properties, in the
Erlangen implementation, or advice on including them in a definition
compatible with Semantic Web technologies if anyone has done this.



Kind regards,

Alli



E-mail: allison.mil...@sysemia.co.uk

Web: www.sysemia.com

-- 


Sysemia Limited

The Innovation Centre, Bristol & Bath Science Park, Dirac Crescent, 
Emerson's Green, Bristol BS16 7FR

Registered in England and Wales. Company Number: 7555456

 

DISCLAIMER

Information contained in this e-mail is intended for the use of the 
addressee only, and is confidential and may also be privileged. If you 
receive this message in error, please advise us immediately. If you are not 
the intended recipient(s), please note that any form of distribution, 
copying or use of this communication or the information in it is strictly 
prohibited and may be unlawful. Attachments to this e-mail may contain 
software viruses which may damage your systems. Sysemia Ltd have taken 
reasonable steps to minimise this risk, but we advise that any attachments 
are virus checked before they are opened.