- From: King . Dany <DKing@drc.com>
- Date: Thu, 25 Jan 2001 15:36:35 -0500
- To: "'www-rdf-logic@w3.org'" <www-rdf-logic@w3.org>
- Cc: TeamXML <TeamXML@drc.com>, "Randolph . Wayne" <WRandolph@drc.com>
- Message-ID: <80090D89EF34D311B69F00508B2C6E4327C1D3@orl01.drc.com>
The following questions/comments are about the DAML+OIL ontology version 1.6 2001/01/11. 1) The "versionInfo" element has a very weak definition: it is only one element with no sub-elements. The use of this element thus far exemplifies the need for a better definition, for example.: <versionInfo>$Id: daml+oil.daml,v 1.6 2001/01/11 20:39:28 mdean Exp $</versionInfo> This element has been overloaded with many types of information: file name, version index, date, time, modifier/creator, and some extra data "Exp". Apparently this string is used by RCS or CVS. Although the use of RCS/CVS is a good practice, the use of this string in an XML document is not. This data should be captured by DAML+OIL in XML fashion: meaningful tags should be made for each type of data. If this string must appear in the document for RCS/CVS purposes, then is should be embedded in an XML comment tag: <!-- -->. However, the versionInfo element should still be populated with the corresponding information. Recommend using DRC Versioning Initiative (DRC-VI), use the following namespace: orl01.drc.com/xml/projects/vi/elements/1.X/drc-ves-1.0v Note: our webpage for the DRC-VI isn't active yet, but will be soon. The element set will contain the following 11 elements: Element: versionInfo --> root element Element: location --> URL of the document Element: replaces --> URL of the previous version Element: replacedBy --> URL of the next version Element: version --> the version identifier Element: email --> email address of the document's creator Element: status --> indicator of the current state of the document, i.e., completed, inwork, published, etc... Element: dc:title --> the name of the document Element: dc:creator --> the name of the creating individual or company Element: dc:contributior --> the name of a contributing individual or company Element: dc:date --> the date of the last status change Hint: the CVS/RCS metadata string can be embedded in the "versionInfo" root element as an XML comment: <!-- -->. 2) Why isn't the "Ontology" element at the beginning of the DAML+OIL ontology? 2.a) Why are there 2 imports: one in the "rdf:Description" element and the other in the "Ontology" element? 2.b) Shouldn't "rdf:Description" element at the beginning of the DAML+OIL ontology be the "Ontology" element? 2.c) Also, the original purpose of the "rdf:Description" element in an ontology was for Mike Dean's crawler. If a robust versionInfo element, such as the one described by the DRC-VI, is used there will be no need for a "rdf:Description" in an otology. All of the necessary information is captured by the DRC-VI element set. 3) Why are there both prefixed and un-prefixed "Property" elements? 3.a) Is there a difference between the two types of definitions? 3.a.i) If so, why? And what is the difference? 3.a.ii) Since the DAML+OIL "Property" element is defined using the "sameClassAs" property to be equivalent to the "rdf:Property", then aren't the prefixed and the un-prefixed uses of "Property" equivalent? 3.a.iii) Or was it just a typo? 4) Since DAML+OIL equivalencies have been defined for RDF and RDFS elements and attributes, why are the prefixed versions used at all? 4.a) Even though the RDF / RDFS prefixed elements and attributes are technically valid in DAML+OIL, shouldn't their use be depreciated? 4.b) And, shouldn't the use of the equivalent DAML+OIL elements be appreciated or strongly recommended? 5) The "Thing" class is not implemented properly. The "Thing" class is supposed to be the base class in DAML+OIL, however, it is currently just another class which has no direct connection to any other class. Per the current DAML+OIL specification, the "rdf:Class" is the base class in DAML+OIL. Please review the attached PowerPoint presentation regarding the problem (each slide has important notes). Also, there are two solutions described therein. The Alternative Solution B is the most concise and is preferred. <<Thing.ppt>> The following questions/comments are directed at the example ontology/instance (mostly v1.4): 6) Can you extend the example ontology and instance about "Adam". For example, include "Eve", "Cain" and "Able", along with both family and non-family oriented relationships. 6.a) How would the slaying of Able by Cain be described? 7) Can you give an example for the "UniqueProperty". Specifically, contrast it with the example already given for "UnambiguousProperty". 8) What does "FvH" mean? It is used in the XML comments throughout the example (v1.1). 9) "hasSpouse" is referred to but never defined. Thank you, Dan King Dynamics Research Corporation Orlando, Florida 407.380.1200x113 dking@drc.com
Attachments
- application/vnd.ms-powerpoint attachment: Thing.ppt
Received on Thursday, 25 January 2001 15:39:59 UTC