RE: Re[2]: Associations in RDF

>Yes , but some notes around an extension of the term "properties" .
>Even in ODMG-93 standard they differ two types of properties as
>"attribute"  type and as "relation" type. From this point of view such
>property of a car as "colour" is "attrbute" type but such property as
>"mark" is "relation" type (we "don't want" for example explicitly
>include in our instance model a relationship with the car manufacturing
>companies). BTW, it is important that a using of monadic and nonmonadic
>predicates is relative (!).
><<<
>
>This is precisely what I am suggesting for RDFS.

Pardon me if this is extremely naive, but what to stop someone defining
subclasses of rdf:Property in their RDF Schema (or would that be
subPropertyOf?), called something like 'Relation' and 'Attribute'?

Cheers,
Danny.