[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [Elist Home]
Subject: Re: [topicmaps-comment] What Is The Type Of A Reified Topic?
* Lars Marius Garshol | | I think what you want is two topics: | | "Web Service using WSDL" | "The WSDL language" * Thomas B. Passin | | Something like this. The real point I'm thinking about is that one | is used for classifying other instances, and the other is the | subject of assertions, remarks, etc. A class topic can perfectly well be the subject of assertions, remarks, and so on, as long as those remarks really pertain to the class. | After all, we can talk about "animal" as if it were an instance of | something (e.g., "an animal can make a good pet") , even though it | is otherwise used as a classification category. Exactly. We can talk about "animal", period. The fact that it may be used as a class is no problem. * Lars Marius Garshol | | I think the relationship is that every single instance of the first | topic makes use of the second topic. That is, you could have a | uses-technology association between every instace of "Web Service | using WSDL" and "the WSDL language". That would be bad modelling for | sure, but I am not sure how to make this assertion about the class. | | Does this work for you? * Thomas B. Passin | | Not yet. If it looks like bad modeling, we probably don't have | quite the right idea yet. Well, what I am saying is that this is bad modelling: [wsdl-service = "Web Service using WSDL"] [wsdl = "The WSDL language"] [currency-calc : wsdl-service = "Bob's currency calculator"] uses-technology(currency-calc : user, wsdl : used) [randomizer : wsdl-service = "Random digit shop"] uses-technology(currency-calc : user, wsdl : used) [amazon : wsdl-service = "Amazon online store"] uses-technology(currency-calc : user, wsdl : used) /* ... more of the same ilk ... */ simply because repeating the uses-technology association for every instance of wsdl-service is redundant; it can be deduced from the fact that the topics are instances of that class. What we need is some kind of existential qualifier that would allow us to say: this assertion holds for every instance of this class. That would look something like this: [wsdl-service = "Web Service using WSDL"] [wsdl = "The WSDL language"] uses-technology(all-wsdl-service-instances : user, wsdl : used) /* need to connect all-wsdl-service-instances & wsdl-service somehow */ [currency-calc : wsdl-service = "Bob's currency calculator"] [randomizer : wsdl-service = "Random digit shop"] [amazon : wsdl-service = "Amazon online store"] /* ... more of the same ilk ... */ Of course, this is kind of an inference rule, and it may be that you want to use that instead. In fact, this highlights the connection between derived topic map information and base topic map information. We haven't really gone into that yet, but we will need to at some point. | What about using scopes? In one scope, the topic participates in a | superclass-subclass association and is the target of instanceOf | values, in another scope it is the subject of various assertions. | Hmm, that would only apply to its name, which is probably not quite | right either. Not sure what you are thinking of here. (Writing it in LTM might help... :-) -- Lars Marius Garshol, Ontopian <URL: http://www.ontopia.net > ISO SC34/WG3, OASIS GeoLang TC <URL: http://www.garshol.priv.no >
[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [Elist Home]
Powered by eList eXpress LLC