- From: Alan Ruttenberg <alanruttenberg@gmail.com>
- Date: Wed, 1 Apr 2009 12:50:37 -0400
- To: W3C OWL Working Group <public-owl-wg@w3.org>
Review of New Features and Rationale. On the positive side, I'd like to pass on a compliment given at a talk I recently gave on OWL, saying that the NF&R was quite helpful. In reviewing the document, the overriding concern I have in reading it is the length. I would like to see it compacted. In my review I give some examples. Detailed review is until section 2.2.2.3 and is adequate to illustrate the general issues and potential solutions. I will continue a detailed review and send further comments. Generally On review I don't see the need for the boxed syntax description. The examples serve to demonstrate enough of the syntax, and typographic emphasis could give an extra indication of which part of the example is demonstrating the feature. In the examples, the subject heading, e.g. the word "HCLS" below "example", don't seem necessary and unnecessarily emphasize the noted bias towards examples from this area. Note: I prefere the expanded "Health Care and Life Sciences" instead of HCLS, in positions such as this, should these be retained. RDF examples need a leading ":" at least, as they need to be namespace qualified. The first example "BrainHemisphere" would be ":BrainHemisphere". There is the need for a careful proofread for grammar. I have given examples where I have seen them. While I support the idea of having RDF syntax examples as well, I would like them to be implemented in the same way as in the other documents, if kept. Mathematic expressions are used to explain some features, but not others. I suggest we don't use them anywhere. Section 2.1 In the boxed examples, link UC#2. This also removes the need to list them as a separate line after the box. Wording (appears for each of the two cases) "Implementations, however, may prefer to take special notices of DisjointClasses for more efficient processing" -> "Implementations, however, may prefer to serialize using DisjointClasses so as to minimize file size and parsing time." "2.1.1 F1: DisjointUnion" use of the addition "F1" for cross referencing doesn't seem necessary. Why not just cite the sections directly? Section 2.1.2 I would remove all or most of this note. If we keep anything, just "The FMA exhibits a huge number of such classes [FMA C in Appendix]" "Note: The FMA exhibits a huge number of such classes [FMA C in Appendix]: 3736 classes of template Left X vs Right X (e.g. Left lung vs Right lung) 13989 classes of template X left Y vs X right Y (e.g. Skin of right breast vs Skin of left breast) 25 classes with template Male X vs Female X (e.g. Male breast vs Female breast) 75 classes X male Y vs X female Y (e.g. Right side of male chest vs Right side of female chest)" If kept, it would be outside the example box. The two examples are virtually the same. Drop one. I like the comment on the second example better: "Nothing can be both a Leftlung and a RightLung" Section 2.1.3 "While OWL 1 provides means to assert values of a property for an individual, asserting that a property has not some values is impossible. This requires the ability to assert facts about an individual stating property values that it does not have" If it were impossible in OWL 1 it wouldn't be syntactic sugar. In fact it is possible but the alternative expression was deemed to be cumbersome and raise the overall difficulty of reasoning because of the use of nominals. It's probably worth writing one version of the OWL 1 way. You also say: "It's possible to take an ontology that is OWL 1 except for NegativePropertyAssertion and preprocess it into an equivalent OWL 1 ontology without it." I think it would be better to have the Theory and Implementation once for the whole section to reduce size, since it applies to all the cases of syntactic sugar. s/Implementaion/Implementation/ Section 2.2.1 Drop: "For example, it might be appropriate to state that the skos:broader relationship is only locally reflexive or, alternatively, locally irreflexive w.r.t. skos:Concept (see also F6). " It's said in F6 anyways and confusing without the larger explanation. "Note: Global reflexivity may be used for local reflexivity in profiles which do not have local reflexivity (e.g., OWL 2 QL)." I don't understand this - global doesn't mean the same thing. Drop. Theory and implementation. "The description logic underlying OWL-DL is SHOIN. OWL 2 is based on a more expressive description logic: SROIQ [SROIQ]. SROIQ extension of SHOIN was designed to provide all possible useful additions to OWL-DL that were requested by users, while not affecting its decidability and practicability." This is something that is globally relevant. Include in the introduction, but not here, I think. The next seems unnecessary, or part of the introduction at best. "SROIQ logic extends SHOIN with reflexive, asymmetric, and irreflexive roles, disjoint roles, a universal role, and constructs ∃ R.Self. It also allows qualified number restrictions and negated role assertions in Aboxes. Additionaly, SROIQ offers complex role inclusion axioms of the form R ◦ S < R or S ◦ R < R to express propagation of one property along another one, which have proven to be very useful in particular for biomedical ontologies (see F8: Property chain inclusion)." Maybe keep this. Interested in feedback from others. But I'd lean towards dropping it. "Local reflexivity is already supported by existing tools, e.g., FACT++. According to developers, local reflexivity was relatively easy to implement [TOOLS] – for any individual x that must have a relationship along a reflexive property, an appropriately labelled edge <x, x> has been added to the model." Section 2.2.2.1 A good example of where dropping the syntax boxes and the domain headers would make this a more easily readable section. Drop "The following examples are some examples of Object Property Cardinality Restrictions from Use Cases among many in HCLS." "Class of objects having at least 5 direct part" s/part/parts/ "Class of objects having atmost 5 Door" -> "Class of objects having at most 5 doors" "Class of objects having exactly 2 RearDoor"-> "Class of objects having exactly 2 rear doors" 2.2.2 Shorten: "As already said above, qualified cardinality restrictions are present in the SROIQ description logic underlying OWL 2 since they were required in various applications e.g.; [Medical Req] [Little Web] and did not pose theoretical or practical problems to be added [SHOIQ]. It has been known from a long time that resulting logic is decidable and QCR was already supported by DAML+OIL, the predecessor of OWL 1. QCRs do not pose an implementation problem either. It has been successfully implemented both in earlier editor, e.g.; OilED, and reasoner, e.g., FACT++, that already processed ontology with QCRs, before OWL 1 recommendation. Current versions of tools under development for OWL 2, e.g.; Protégé 4, FACT++, PELLET, RACER, KAON2 also deals with QCRs [TOOLS] [OWL API]." "It has been known for some time that OWL extended by qualified cardinality restrictions is decidable and they were already supported by DAML+OIL, the predecessor of OWL 1[cite]." 2.2.3 I suggest we remove this, as we don't support the antisymmetric properties needed for mereology, and in any case you can't use these property types with the transitive (i.e. composite) property part_of. "For example, in mereology, the partOf relation is defined to be transitive (if x is a part of y and y is a part of z, then x is a part of z), reflexive (every object is a part of itself), and antisymmetric (if an object has a part which in turn has part itself, then they are the same). Many applications, particularly those where it is necessary to describe complex structures, such as life science applications or more generally for systems engineering, require extensive use of part-whole relations, axiomatized according to these principles" Also drop "Other relations encountered in ontology modeling require similar axiomatizations as well, possibly with different sets of characteristics (see, e.g., [OBO] [RO])." "Another interesting example" -> "An interesting example" "SKOS specification" -> "The SKOS specification" Rewriting a bit: "The SKOS specification [SKOS] makes no committment to the reflexivity or irreflexivity of the skos:broader relationship. The specification mentions that skos:broader should be considered reflexive "for a direct translation of an inferred OWL subclass hierarchy", but that irreflexive would be "more appropriate for most thesauri or classification schemes". With the addition of reflexivity and irreflexivity, OWL 2 allows one to express either of these choices. Self restrictions are even more fine grained, by allowing one to state that skos:broader should only be locally reflexive or irreflexive w.r.t. skos:Concept." Move to example box: One can make skos:broader locally irreflexive w.r.t. skos:Concept via an axiom SubClassOf(skos:Concept ComplementOf(ObjectHasSelf(skos:broader))) i.e. if x is a skos:Concept, then NOT( x skos:broader x ). Section 2.2.3.1 I don't think that we need the mathematical expression. We don't do that uniformly. Comment says: "Everybody has the same blood group as himself" Should say "Every thing has the same blood group as itself". If you want to make it about people then it needs to be locally reflexive. 2.2.3.3 I believe the example violates the global constraints. AsymmetricObjectProperty( proper_part_of ) As propert_part_of is transitive. Theory and Implementation Include only "see F4 Theory and Implementation above Impementation was similar to the approach taken to support local reflexivity."
Received on Wednesday, 1 April 2009 16:51:28 UTC