Skip to content
vangelisv edited this page Sep 13, 2010 · 1 revision

Status – Issues

alpha/pre-alpha.

  • owl2_model – core axiom and description predicates are complete. See below for open issues
  • owl2_from_rdf – incomplete
    • rename?
    • incomplete. TODO:
      • axiom annotations
      • other axioms?
  • owl2_java_owlapi – highly incomplete
  • owl2_xml_parser – nearly complete. Targeted at old P4 output, needs amended to correspond to spec.
  • swrl – reasonably complete
  • reasoners – incomplete

To be ported from Thea1:

  • DIG
  • Grosof mapping

OWL2 Model Open Issues

Structure Sharing

Should we allow bnode IDs as arguments of predicate axioms for expressions?

Enumeration of expressions

We provide semi-deterministic predicates of the form
?type(+Expression). Should the mode be extended to allow
enumeration of all descriptions/expressions? This would probably
require forcing all expressions to be bnodes OR possibly recursively
analyzing the term Axiom in a call axiom(Axiom)

(note: I have added axiom_references/2 and axiom_indirectly_references/2 – these could be used to iterate over all referenced descriptions)

Type checking

Is Tom Schrijvers type checking system going to be integrated into SWI and Yap? Should we use that?

I am attempting to put as much typing info in the pldoc comments,
but unsure of the conventions for complex terms

Ontologies

continue using ontologyAxiom/2? Alternatively use builtin prolog module mechanism..?

Axiom Annotations

I’m not sure about the proposed model for axiom annotations. Should’nt we include the as (optional) argument
in the axiom it self? ie subclassOf(AList, C1,C2) plus a rule subclassOf(C1,C2) :- subclassOf([],C1,C2). ???

Other Open Issues

how does this fit into SWI SemWeb stack

Porting

  • some parts work with Yap
  • XSB compat will be more of a challenge
  • Other prologs?

Yield prolog?