- From: Paul Wilton <paul.wilton@ontoba.com>
- Date: Fri, 9 Sep 2011 10:20:05 +0100
- To: semantic-web@w3.org
- Cc: "jarred.mcginnis" <Jarred.McGinnis@pressassociation.com>, Bernard Vatant <bernard.vatant@mondeca.com>
Hi Bob In most of these cases you have pointed out, we have inherited from the classes/properties you mention for very good practical reasons, either as we want to specialise further the class in question (eg sub_event on the public domain event ontology was not transitive, so we have defined our own specialized property that is transitive). The other very practical reason is that we *want* to define a ontology for news with its own classes and predicates. We have subclassed public domain ontologies where we feel the public domain ontology is widely used, so that when RDF is published we wil also be publishing the ancestor classes and properties for consumers that understand it. This also addresses the practical architectural problem of contract binding. Developers of APIs built around this ontology can bind to the press.net classes and properties, decoupling themselves from public domain ontologies which our out of their control. As these (public domain) ontologies evolve the press.net ontology can be modified accordingly with respect to the public domain ontologies while not breaking the contracts of those clients bound to the press.net ontology. This is good practice in my opinion. We are considering dropping the skos inheritance altogether, as skos is particularly cumbersome around inverseOf properties and their own ancestors resulting in n*n*8 statement chains being inferred from a n-depth classification scheme - (using pns:subClassificationOf subPropertyOf skos:narrower) best regards Paul From: Bob Ferris <zazi@smiy.org> Date: Thu, 08 Sep 2011 20:00:26 +0200 Message-ID: <4E6902BA.10207@smiy.org> To: semantic-web@w3.org Hello again, ... and I can continue the previous list: 14. pnt:about can be replaced by foaf:topic/foaf:primaryTopic. 15. pnc:classifiedBy can be replaced by dcterms:subject (as recommended by SKOS). 16. pnc:Classification can be replaced by skos:Concept. 17. pnc:subClassficationOf can be replaced by skos:broader. 18. pns:aliad can be replaced by skos:altLabel. 19. pns:Stuff can be replaced by owl:Thing. So far the following terms are rather well designed: 1. pns:mentions 2. pns:notablyAssociatedWith Please keep in mind: 1. Ontology modelling is not database schema modelling 2. Ontology modelling is not MDA modelling 3. Ontology modelling should be intended to create a kind of 'shared meaning' or 'shared understanding', i.e., - try to reutilize as much as possible (its an open, shared knowledge space) - think carefully about introducing new terms (check their definitions against definitions of existing terms) Although at a first glance it seems to be practical to have many terms in one's own namespace. However, it will cause more difficulties when should utilise a dataset that is created with these ontologies (yes we can do ontology alignment, but this is more expensive than creating a 'shared meaning' direclty). Then we do not have much differences between existing database and API approaches, where everyone defines one's own knowledge space. Our goal is to create a shared knowledge space like we have shared languages, e.g., English, or? Nevertheless, the freedom of the open knowledge space is that everyone can say everything about everything, and, hence, define as many new ontologies as one like to create. Whether this is practical or not, is another issue ... Cheers, Bo PS: I'm not seeking for the one and only gold standard ontology space ;) On 9/8/2011 6:35 PM, Bob Ferris wrote: > Hi Jarred, > > at a first glance, here are my remarks: > > 1. pne:Event, pne:sub_event seem to be a bit duplicated. I guess, > event:Event, event:sub_event are enough. > > 2. pne:title can be replaced by, e.g., dc:title. > > 3. pns:Person can be replaced by foaf:Person. > > 4. pns:Organization can be replaced by foaf:Organization. > > 5. pns:worksFor can be replaced by rel:employedBy [1]. > > 6. pns:Lcoation can be replaced by geo:SpatialThing > > 7. Re. the tagging terms, I would recommend to have a look at the Tag > Ontology [2] or similar (see, e.g., [3]) > > 8. Re. biographical events I would recommend to have a look at the Bio > Vocabulary [4], e.g., bio:birth/bio:death. > > 9. pns:label can be replaced by dc:title (or rdfs:label). > > 10. pns:comment can be replaced by dc:description (or rdfs:comment). > > 11. pns:describedBy can be replaced by wdrs:describedby [5]. > > 12. Re. bibliographic terms I would recommend to have a look at the Bibo > Ontology [6], e.g., bibo:Image (or foaf:Image), or the FRBR Vocabulary > [7], e.g., frbr:Text. > > 13. pna:hasThumbnail can be replaced by foaf:thumbnail. > > ... > > Please help us to create 'shared understanding' by reutilising terms of > existing Semantic Web ontologies. > > Cheers, > > > Bo > > > [1] http://purl.org/vocab/relationship/employedBy > [2] http://www.holygoat.co.uk/projects/tags/ > [3] > http://answers.semanticweb.com/questions/1566/ontologyvocabulary-and-design-patterns-for-tags-and-tagged-data > > [4] http://purl.org/vocab/bio/0.1/ > [5] http://www.w3.org/2007/05/powder-s#describedby > [6] http://purl.org/ontology/bibo/ > [7] http://purl.org/vocab/frbr/core# > > On 9/8/2011 3:48 PM, Jarred McGinnis wrote: >> Hello all, >> >> The Press Association has just published our first draft of a 'news' >> ontology (_http://data.press.net/ontology_). For each of the ontologies >> documented, we've included the motivation for the ontologies as well as >> some of the design decisions behind it. Also, you can get the rdf or ttl >> by adding the extension. For example, >> http://data.press.net/ontology/asset.rdf<http://data.press.net/ontology/asset.rdf>gives >> >> you the ontology described at http://data.press.net/ontology/asset/ .. >> >> Have a look at the ontology and tell us what you think. We think it is >> pretty good but feel free to point out our mistakes. We will fix it. Ask >> why we did it one way and not another. We will give you an answer. >> >> Paul Wilton of Ontoba has been working with us at the PA and has spelled >> out a lot of the guiding principles of this work at >> http://www.ontoba.com/blog. >> >> The reasons behind this work were talked about at SemTech 2011 San >> Fransisco: >> http://semtech2011.semanticweb.com/sessionPop.cfm?confid=62&proposalid=4134 >> >> <http://semtech2011.semanticweb.com/sessionPop.cfm?confid=62&proposalid=4134> >> >> >> Looking forward to hearing from you, >> >> *Jarred McGinnis, PhD*
Received on Friday, 9 September 2011 12:31:31 UTC