Misleading description of org:OrganizationalUnit

Published on: 16/11/2016

We have described the class org:OrganizationalUnit as a subclass of cpov:PublicOrganization. We have done so implicitly in the text and declaratively in the diagram.

In the text we write:
"An Organizational Unit is a sub class of Public Organization [...]"
..which of course can be true and does not in itself declare OrganizationalUnit as subclass of PublicOrganization, but when the reader looks at the UML diagram there is a generalization between the OrganizationalUnit and the PublicOrganization. The error is clearly in the diagram but is is not unlikely that the reader will see the textual description as a confirmation  of the information given in the diagram.

I propose that we make the textual description more precise and that we move the generalization symbol in diagram to instead show that org:OrganizationalUnit is a subclass of org:Organization as defined in Organizational Ontology.

Peter Bruhn Andersen
Agency of Digitisation






Wed, 16/11/2016 - 16:55

Although the quotation from the document is confusing (I guess the author had “Organization” instead of “Public Organization” in mind) I agree with the proposal: an OrganizationalUnit is an Organization, but is not a Public Organization, since (as mentioned implicitly in the document) it has no legal status. Just moving the subclass link from PublicOrganization to Organization solves the issue only partially, however, since it does not imply that OrganizationaUnit and PublicOrganization are disjoint classes (in principle, a disjoint constraint should be introduced). In addition, I believe it would be preferrable to include a composition link between PublicOrganization and OrganizationalUnit (a PublicOrganization may be composed of two or more OrganizationalUnits).

philarcher (not verified)
Fri, 25/11/2016 - 12:15

I have amended the diagram and text to show Organizational Unit as a sub class of Organization and amended the text accordingly, thus fixing my error. Thank you. It is clear that an Organization may comprixse several sub organizations, I don't think there's any more to be done there. Since we're reusing the ORG ontology, I don't want to assert disjointedness beyind what is already declared at the bottom of https://www.w3.org/ns/org.