http://ontologydesignpatterns.org/index.php?action=history&feed=atom&title=Reviews%3AStefanoDavid_about_OnlynessIsLoneliness_%28OIL%29 2026-04-29T23:58:09Z Revision history for this page on the wiki MediaWiki 1.25.6 http://ontologydesignpatterns.org/index.php?title=Reviews:StefanoDavid_about_OnlynessIsLoneliness_(OIL)&diff=5794&oldid=prev 2009-09-10T14:07:16Z

<p>New page: {{Content OP Proposal Review Template |CreationDate=2009/9/10 |SubmittedBy=StefanoDavid |ContentOPUnderReview=OnlynessIsLoneliness (OIL) |RevisionID=5787 |Score=1 - needs minor revision |R...</p> <p><b>New page</b></p><div>{{Content OP Proposal Review Template<br /> |CreationDate=2009/9/10<br /> |SubmittedBy=StefanoDavid<br /> |ContentOPUnderReview=OnlynessIsLoneliness (OIL)<br /> |RevisionID=5787<br /> |Score=1 - needs minor revision<br /> |ReviewSummary=The authors propose a solution to a common modeling error, the use of disjointness assertions, that causes inconsistencies. <br /> |ReviewConfidence=Good knowledge in Description logics and OWL<br /> |ReviewProblems=* I would rephrase the problem statement: I would suggest to call &quot;problem&quot; the modeling error (which is obviously an Anti-Pattern, as it causes the ontology to become useless), and &quot;pattern&quot; the proposed solution.<br /> <br /> * There are different hydrontologies in the &quot;Pattern solution example&quot;: I would point only to one of them, as they seem very similar, and use it also as use case.<br /> <br /> * Although the hydrontologies seem to be some official Spanish knowledge bases, I would rather use an english version of a part of them that clearly states the problem, as it is not so comfortable to search in the whole ontology the pattern.<br /> <br /> * I would extend the definition of this pattern by suggesting that it is allowed the use of multiple restrictions (i.e., C1 subClassOf R only C2, ... ,C1 subClassOf R only Cn) to be combined in a single disjointness axiom, e.g., for those languages, like OWL2, that allow DisjointClasses into a single axiom:<br /> C1 subClassOf R only (C2 or ... or Cn); disjointClasses (C2, ... Cn)<br /> <br /> * &quot;We have categorized them into three groups:&quot;. There are no three groups in the remainder, but the description of three classes<br /> <br /> <br /> |ReviewRelevance=Medium/High, since the wrong use and the misuse of disjointness is often difficult to spot.<br /> |ReviewBestPractice=High, as it solves modeling mistakes.<br /> |ReviewReusability=Yes, across different languages and every time a (wrongly defined) disjointness is involved<br /> |ReviewUnderstandability=The modeling problem and proposed solutions need a rewording, but the purpose is clear.<br /> |ReviewClearProblem=No, it needs some rework. <br /> |ReviewClearRelevance=yes<br /> |ReviewFigures=None provided<br /> |ReviewMissing=ER/UML diagram<br /> }}</div>

StefanoDavid