0

Actually, I'm getting lost about what can be done and cannot be done using a reasoner in general.

My question is two fold:

  1. I used to think that a reasoner is only dedicated for sumbsumption and taxonomies. Recently, I discovered that a reasoner can be used in multiple ways within some algorithms for example this reference introduces using a reasoner to search for entities compliant to a specific pattern in the ontology:

    • What is meant by a DL generic reasoning algorithm and how can it be implemented through a reasoner?

    • Can you please refer my to any references where I can get more familiar to things that can be achieved by reasoning algorithms if any?

  2. My original problem is that: given an OWL ontology and two classes C1 and C2, I want to check whether there is a functional property chain connecting them, i.e. let's consider (C1 P1 C3), (C3 P2 C4)and (C4 P3 C2) where P1, P2 and P3 are functional object properties, thus the result will be (P1, P2, P3) which is the functional property chain connecting C1 and C2.

Median Hilal
  • 1,483
  • 9
  • 17
  • "a reasoner can be used in multiple ways" Such as what? (I see that you've provided a link, but please summarize here.) Many reasoning tasks can be expressed subsumption. E.g., asking whether p(a,b) is the same as asking whether "{a} ⊑ ∃p.{b}". – Joshua Taylor Nov 11 '14 at 12:31
  • I quote here about this specific algorithm "This algorithm was conceived to match concepts against patterns to support the pruning of large concept descriptions discarding unimportant aspects. More precisely, given a concept pattern D (i.e., a concept description containing variables) and a concept description C without variables, the matching problem asks for a substitution σ (of the variables by concept descriptions) such that C is subsumed by σ(D). – Median Hilal Nov 11 '14 at 21:32
  • For instance, considering the pattern “the research interest of those people being only interested in a single research area” (expressed in DL as ∀research-interests.X), and the concept description “those people only having cats as pets and being only interested in AI research” (∀pets.Cat ∩ ∀research-interests.AI), the matching algorithm finds the scientific interests (in this case Artificial Intelligence) described in the concept (i.e., assigns to X the value AI)." – Median Hilal Nov 11 '14 at 21:33

0 Answers0