- From: Deborah L. McGuinness <dlm@ksl.stanford.edu>
- Date: Wed, 13 May 2009 19:44:33 -0400
- To: W3C OWL Working Group <public-owl-wg@w3.org>
Below please find my review of the OWL2 Primer. Note that I am willing to re-review sections 9 and 10 where I think a significant rewrite needs to happen. Deborah Review of OWL 2 Primer http://www.w3.org/TR/2009/WD-owl2-primer-20090421/ and the wiki version http://www.w3.org/2007/OWL/wiki/Primer Working draft April 21 2009 Note, I did not check all of the syntaxes and in fact, I only reviewed the examples in the functional style syntax. I assume the examples have been checked for syntactic correctness. 1. Summary of Changes – I understand that this is great to have in now but it should come out before final publication. 2. On: “The key goal of the primer is to help develop insight into OWL, its strengths, and its weaknesses. The core of the primer is an introduction to most of the language features of OWL by way of a running example. Most of the examples in the primer are taken from a sample ontology (which is presented entirely in an appendix <http://www.w3.org/TR/2009/WD-owl2-primer-20090421/#The_Complete_Sample_Ontology>). This sample ontology is designed to touch the key language features of OWL in an understandable way and not, in itself, to be a example of a good ontology.” 1. After we mention “most of the language features” add a sentence saying for a listing of language features, see the Quick Reference document which then provides links into the appropriate sections of the appropriate documents concerning syntax and examples. 2. Also, change “a example” to “an example” 3. On “In Section 9 <http://www.w3.org/TR/2009/WD-owl2-primer-20090421/#OWL_2_DL_and_OWL_2_Full> we address the differences between OWL 2 DL and OWL 2 Full,” 1. Owl dl and full have not been introduced. We need to say something like the two species of owl but I realize we got rid of the use of species. What I am not sure of is what has replaced it. We need to say something even if it is only two views of OWL 2 (I notice later we mention that there are 2 ways of thinking of owl…) 4. On “while in Section 10 <http://www.w3.org/TR/2009/WD-owl2-primer-20090421/#OWL_2_Profiles> we describe the three Profiles of OWL 2”. 1. Suggest something like “while in Section 10 <http://www.w3.org/TR/2009/WD-owl2-primer-20090421/#OWL_2_Profiles> we describe the three Profiles of OWL 2, three sublanguages of OWL”. 5. On “For readers already familiar with OWL 1, [OWL 2 New Features and Rationale <http://www.w3.org/2007/OWL/wiki/Primer#ref-owl-2-new-features>] provides a comprehensive overview of what has changed in OWL 2.” 1. Change to “For readers already familiar with OWL 1, [OWL 2 New Features and Rationale <http://www.w3.org/2007/OWL/wiki/Primer#ref-owl-2-new-features>] provides a comprehensive overview of what has changed in OWL 2. This document also replaces the OWL Guide <http://www.w3.org/TR/2003/CR-owl-guide-20030818/> provided for OWL 1. 6. On “OWL is part of the Semantic Web, so…” 1. This feels awkward to me, I would say something like “OWL is a language used to describe semantic web terms, so …” 7. On “ … all four syntaxes “ 1. Change to “… all five syntaxes” 8. On “Only if a specific implementation complies with these semantics, it will be regarded OWL 2 conformant” 1. Change to “Only if a specific implementation complies with these semantics, will it be regarded OWL 2 conformant” or better “Only implementations that comply with these semantics will be regarded as OWL 2 conformant” 9. I find the following open to confusion “In particular, there is no way to enforce that a certain piece of information (like the social security number of a person) has to be syntactically present.” 1. One reading of that for me is that min cardinality restrictions are not there. What I believe this is really aiming at is the open world assumption so that just because a ss# is not there now, does not mean it might not be there later. 2. This paragraph and the next one basically on what owl is not I think belong buried later – not so far near the beginning where I think they have more potential to cause confusion than to help. 10. On “OWL can rather be considered…” 1. I would change to “OWL can be considered ….” (the previous form sounds a bit like a translation from German, while correct it sounds better without rather to me. 11. Concerning …” a particular OWL ontology is true.” 1. I usually do not think of an ontology as being true but rather that a set of statements that together form an ontology are true. The sentence is also a little awkward with the introduction of interpretation. So I would change this to “The formal semantics of OWL specifies, in essence, for which possible “states of affairs” – a particular set of OWL statements is true. 12. Concerning “It is a strength because OWL 2 can discover…” 1. OWL 2 is not the thing discovering information, it is the tools supporting OWL 2 doing the discovery. I would change to “It is a strength because OWL 2 tools can discover…” 13. On the paragraph starting with “One can use basic algebra….” 1. This feels out of place – it is a more sophisticated notion than most of the rest of the writing and is only for a subset of users. It should somehow be noted that most will want to skip this. It could be a (granted long) footnote. 14. On “Thereby we will represent information about a particular family. (We do not intend this example to be representative of the sorts of domains OWL should be used for, or as a canonical example of good modeling with OWL, or a correct representation of the rather complex, shifting, and culturally dependent domain of families. Instead, we intend it to be a rather simple exhibition of various features of OWL.)” 1. I do not think the parenthetical adds a lot but it does detract. This is now the third place where it seemed that a thought was more appropriate for a footnote if it was to be kept. 15. On “We first need to provide the information what persons we are talking about. This can be done as follows:” 1. This sentence is awkward. Something like the following may be better: A few persons need to be created…. 16. On section 4.5 property hierarchies, the text starting with “There is also a syntactic shortcut…. goes beyond the right side of the box. I presume this is font size dependent. Is there a way to make it stayin the box? 17. In section 4.6, I would give a forward pointer to local value restrictions from range. I would just say something like there is a related construct and then provide a link. 18. In 4.7, we might strengthen the point about the unique names assumption something like the following . Instead of 1. “This “unique names assumption” would be particularly dangerous in the Semantic Web, where names may be coined by different organizations at different times unknowingly referring to the same individual” try 2. This lack of a required “unique names assumption” is particularly well suited to Semantic Web applications where names may be coined by different organizations at different times unknowingly referring to the same individual 19. In 4.7, I suggest a different example for the sameIndividual something like James and Jim or Deborah, Debbie, and Deb if we want more. 20. 5.2 starts awkwardly. 1. Suggest changing “By property restrictions we understand another type of logic-based constructors for complex classes. As the name suggests, property restrictions are constructors involving properties. The first property restriction called /…./ 2. /To/ 3. Property restrictions provide another type of logic-based construction for complex classes. As the name suggests, property restrictions use constructors involving properties. One property restriction called /….”/ 21. In the happy child example, it sounds like the only way you can guarantee that you have a happy child is with some values from. You could also do it just by using a minimum cardinality restriction on hasChild in combination with the allValuesFrom Happy person. 22. On the john’s 4 children who are parents example, it sounds odd to refer to parents and children as that and which when we expect them to be people. I would use who. Thus change “arbitrarily many further children which are not parents.” To “arbitrarily many further children who are not parents.” And change “at least two children that are parents:” to “at least two children who are parents” 23. On 5.4, I would change “Therefore, classes defined this way are sometimes referred to as /closed classes/.” To “Therefore, classes defined this way are sometimes referred to as /closed classes/ or enumerated sets”. 24. In 6.1, after the introduction of the functional property hasHusband, I would include that using functional properties, a reasoner can infer that if for example we have a statement that my husband is James and another that my husband is Jim, then Jim and James must refer to the same individual. 25. In section 7, wow – 200 is a seriously large max for human age. I would drop it to at least 150. 26. In section 8.1 (in the wiki) I see the editor comment about replacing rdfs:label with rdfs:comment but I do not see this in the example syntax. We might also add that often such comments are used in interfaces to provide access to natural language text to be displayed in help interfaces. 27. On section 9, I agree with the comment by mike smith on may 13 that this section could use some rework. The intro sentence of 2 ways of thinking about owl 2 seems odd to me as well. I am willing to re-review when the updated 9 is in. 28. On section 10, I am not sure what goes in this section and what goes in the owl profiles document. It seems like the main thing I would want to get on profiles from this document is a sentence or 2 on each profile and why one chooses that profile and a small example. The current version seems to have too much content and I agree with mike that complexity class, links to literature, and history do not belong in this document. I am also willing to re-review when the update to 10 is in. 29. On section 12, what to read next, I would also include a link to the quick reference to have a listing of constructors along with hyperlinks into relevant documents. 30. References should be updated to include any new references including the quick ref, the owl 1 guide this is replacing
Received on Wednesday, 13 May 2009 23:44:51 UTC