OWL2 annotations

Following further study of the OWL2 documents, I have a question about
extended annotations. 

 

Is the following permitted (e.g. in Manchester Syntax)

 

Class: AcuteHeartDisease

        EquivalentTo: Disease 

                and (has_location some Heart) Annotations: layer general,

                and (has_feature some (Onset that (has_state some acute)
Annotations: layer possible)) Annotations: layer possible

 

Class: AcuteLiverDisease

        EquivalentTo: Disease

                and (has_location some Liver) Annotations: layer possible,

                and (has_feature some (Onset that (has_state some acute)
Annotations: layer possible)) Annotations: layer possible

 

Class: FemoralFracture

         EquivalentTo: Fracture

                that (has_location some Femur), Annotations: layer possible

 

The purpose of the annotations in these examples is to allow the editing of
an ontology as a whole to be continuously constrained by a layered set of
ontological commitments, such as was used extensively in the OpenGALEN
project. A similar approach is emerging as one of the mechanisms to
implement ontological commitments within SNOMED CT.

 

For example, consider that you may want to apply the following semantic
'rules':

 

Disease generally has_location Anatomy

Necrosis possibly has_location Heart

 

.such that AcuteHeartDisease and FemoralFracture (above) should be permitted
to be reified within the ontology, but AcuteLiverDisease (as defined AND
constrained above) should not, whereas defined and constrained as below, it
should be permitted:

 

Class: AcuteLiverDisease

        EquivalentTo: Disease

                and (has_location some Liver) Annotations: layer general,

                and (has_feature some (Onset that (has_state some acute)
Annotations: layer possible)) Annotations: layer possible

 

OWL2 annotations appear to allow representation of the set of ontological
commitments as annotations, probably on the ontology as a whole though also
possibly on the individual domain classes for each commitment statement.
However, I have a suspicion that there isn't a mechanism to tune *which*
layers, in a layered model of ontological commitments, are to be applied at
the level of individual restrictions within individual EquivalentTo or
SubClassOf frames of individual class definitions.

 

In GALEN, the tuning was done using two distinct operators ('which' and
'whichG'):

 

(PathologicalBodyProcess 

whichG <hasUniqueAssociatedProcess ((NeoplasticProcess which hasMalignancy
(Malignancy which hasAbsoluteState malignant))

which actsSpecificallyOn GranulocyteStemCell)>) 

which hasSyndromeElement ((CountConcentration 

which isCountConcentrationOf (Monocyte 

which isInSuspensionWithin (Blood which hasPhysicalState  (PhysicalState
which hasAbsoluteState liquid)))) 

which hasQuantity (Level which hasMagnitude highLevel))

name MonocyticLeukaemia

 

These appear, at least superficially, to correspond to different flavours of
the 'someValuesFrom/that/and' construct in OWL. This leads me to suspect
that - if the flavours were instead to be signified as OWL annotations -
you'd need something similar to my examples above.

 

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

Dr Jeremy Rogers MD MRCGP DRCOG DFFP MB ChB

Principal Terminology Specialist

Technology Office (Leeds)

NHS Connecting for Health

 

jeremy.rogers@nhs.net

+44 7811 525 314

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

 


***********************************************************************
This  message  may  contain  confidential and  privileged  information.
If you  are not the  intended recipient  you should not  disclose, copy
or distribute information in this e-mail or take any action in reliance
on its contents.  To do so is strictly  prohibited and may be unlawful.
Please  inform  the  sender that  this  message has  gone astray before
deleting it.  Thank you.

2008 marks the 60th anniversary of the NHS.  It's an opportunity to pay
tribute to the NHS staff and volunteers who help shape the service, and
celebrate their achievements.

If you work for the NHS  and  would like  an NHSmail  email account, go
to: www.connectingforhealth.nhs.uk/nhsmail
***********************************************************************