Re: [OEP] Inference Patterns

Hi Mike, yours is a major need for actual ontology deployment. 
Essentially, you want to define relation compositions, either for 
introducing a new relation, or to make a view (query). Since 
properties cannot be defined in OWL (only primitive characterization 
is allowed), to do what you need either you use SWRL, or (if the 
domain allows), you could try with multihierarchical classification, 
e.g.:

Class(ReportSubject complete
  Subject
  (restriction subjectOf someValuesFrom
   (intersectionOf
     Conference
     (restriction aboutnessOf someValuesFrom Report))))

then you can look for report subjects, and look through which 
conferences/subjects they are subject of.
Still another option is using embedded queries.

More at next, I have to go now.
Ciao
Aldo

At 15:06 -0700 21-10-2005, Uschold, Michael F wrote:
>Here is a simple inference that I wish to get my OWL ontology to do for
>me, and I don't know whether it is possible.
>There might be a nice trick, but I don't know.  This is one of many
>commonly arising examples of wanting certain kinds of inferences to go
>throug, and could be the subject of a note.
>
>Here is my example:
>
>I have an ontology which is mainly about tracking reports of people's
>activities in an organization (e.g. publications, attending conferences,
>workshops, making customer contacts, etc).
>
>Key classes include Report, Conference, Publication, Subject etc.
>
>The key relation is: isAbout which says what the Report is about.
>e.g.  isAbout(Report1, ISWC-05) and/or isAbout(Report1, Publication23).
>One report could be about more than one thing.
>
>We also have a relation indicating the subject of
>1) the things that Reports are about (a conference), and
>2) the Reports themselves.
>
>I WANT TO INFER THE SUBJECT OF THE REPORT FROM THE SUBJECT OF THE THINGS
>THAT THE REPORT IS ABOUT. 
>
>For example:
>
>isAbout(?Report, ?Conference) &
>hasSubject(?Conference, ?Subject)  ==> hasSubject(?Report,?Subject)
>
>
>This message has two purposes:
>1. how can I achieve this particular task, if at all
>2. there are LOTS of such examples where one wishes to get an inference
>to go through, and it is not always easy to know how or whether it can
>be done. Should we have a note to address this kind of thing?
>
>============================================
>Mike Uschold
>Mathematics and Computing Technology -- Phantom Works;
>Tel: 425 865-3605              Fax: 425 865-2965
>Building: 3307 Cube: 33D3     
>============================================


-- 



Aldo Gangemi
Research Scientist
Laboratory for Applied Ontology
Institute for Cognitive Sciences and Technology
National Research Council (ISTC-CNR)
Via Nomentana 56, 00161, Roma, Italy
Tel: +390644161535
Fax: +390644161513
aldo.gangemi@istc.cnr.it
http://www.istc.cnr.it/createhtml.php?nbr=71

Received on Monday, 24 October 2005 12:28:16 UTC