A search for tools
Yesterday and today have been taken up with a very quick look at a few tools to play with rdf. I started with cwm, written a while ago. Checks, converts from one format to another.. and maybe more. Next came pellet which again is a verification tool. I'd seen (but not understood) isaviz, a graphical tool (I think primarily intended for generation) which I used today to import an N3 owl file. Neat. Finally, today I installed one of Dave Becketts redland tools, rdfproc and rapper. I had a look at rdflib - a python tool; but its lack of documentation I found offputting.
I'm starting to think better too - tools and syntax aren't getting in the way quite so much. I'm hitting one problem quite hard. When designing an ontology, my only previous experience is with OO modelling and schema (DTD and Relax NG, not rdf). I'm finding that thinking in graphs whilst talking about classes is not simple. E.g. I have a book class. I want people to tell me what the dc:subject property of the book is. Now that is clearly a property on an instance (sorry, individual), not on the class itself. Is that how I specify it? Or do I simply list the property in the ontology so that people can use it? I don't know. More reading.
Anyway, I'm gathering tools which should help. Using them efficiently is another problem. The result? My map of rdfland has grown in size, but not in comprehension.
Amendment. A request on the right irc channel brings back the fact that rdflib isn't built with the documentation, its separate (for the api documentation) and even an article on xml.com. Seems like some useful reading, and I prefer Python.
I guess I should add seasonal greetings to both my readers. May your $DEITY be with you, and best wishes for 2006.
Keywords: rdfComments (View)
Return to main index