On 20 May 2011 11:43, Egon Willighagen <egon.willighagen at gmail.com> wrote:

> On Fri, May 20, 2011 at 10:19 AM, Nina Jeliazkova
> <jeliazkova.nina at gmail.com> wrote:
> > Related to the recent discussion on RDF scalability, I am sending an
> example
> > how simple use of prefixes in RDF can reduce the size of the serialized
> RDF.
> >  The length of the newly generated prefixed files  (both in RDF/XML and
> > N3)  is *3 times less* than the original one (RDF/XML).
>
> A further note is on:
>
> http://apps.ideaconsult.net:8080/ambit2/compound/9908/conformer/534823
>
> Now, namespaced resources are not happy (at least not in Jena) about
> '/'s in resources,


Indeed.


> so you cannot abbreviate this into:
>
> ambit:compound/9908/conformer/534823
>
> But, it might be worthwhile to use a few NSs, so to give:
>
> amconf:534823
>
> with
>
> @prefix amconf:
> <http://apps.ideaconsult.net:8080/ambit2/compound/9908/conformer/>
>


It might not be a big win, if these are not repeated.

But what I am wondering is why having

@prefix otm:     <http://webservices.in-silico.ch/model> .


doesn't help to have

ot:hasSource otm:124 ;

instead of

ot:hasSource <http://webservices.in-silico.ch/model/124> ;

Any ideas?  The model URI is repeating many times, it will be quite a gain
if it is prefixed.


>
> But that a valid QName :(
>
> In that respect, it would be good for a future OpenTox API to have the
> last bit a valid NCName, so starting with a 'a-Z' or '_':
>
> http://www.w3.org/TR/REC-xml/#NT-NameStartChar


Hm, does this mean it's actually the numbers that are messing up with the
 RDF prefixes ?  It will be trivial to add an alphabetical first char to any
of the URIs.


There are other things besides prefixes, to be optimized in this particular
example, but I'll send my thoughts in another mail.

Nina


>
>
> Grtz,
>
> Egon
>
> --
> Dr E.L. Willighagen
> Postdoctoral Researcher
> Institutet för miljömedicin
> Karolinska Institutet (http://ki.se/imm)
> Homepage: http://egonw.github.com/
> LinkedIn: http://se.linkedin.com/in/egonw
> Blog: http://chem-bla-ics.blogspot.com/
> PubList: http://www.citeulike.org/user/egonw/tag/papers
> _______________________________________________
> Development mailing list
> Development at opentox.org
> http://www.opentox.org/mailman/listinfo/development
>