> I�m am following the mappa mailing list and since there is some recent
> traffic today, I thought I post my questions ;-)
:)
> 1)
> Would it make sense to convert a topic map to a google protocol buffer
> for a rest service? There is a implementation present in python, java
> and I think perl too. The size is supposed to be at least twice as
> small as json and the data is faster processed.
If the resulting topic map is reasonable small, it could make sense
(IIRC the protocol buffer msgs shouldn't be bigger than 1MB).
I don't know if the saved space and maybe the improved processing
performance would make a big/notable difference, though.
> 2)
> Is there a way to create a large topic map from dbpedia data? This
> would be very handy for performance testing. Like a 100MB, 1GB and 4GB
> topic map file.
Yes, this should be doable. Not via Mappify, though (Mappify's
responses are limited to 1MB due to restrictions of GAE). But it
should be very easy to generate topic maps in different sizes from the
DBpedia dump offline.
> Is there already a dbpedia-topic-map online and
> downloadable. If not can I create one using mappa and mappify?
DBpedia publishes RDF files and Mappa cannot read RDF yet. But you can
use MIO [1] to create topic maps from RDF sources. MIO is also used by
Mappify.
The TMLab created MapMark [2] but it is not published yet. At least I
cannot find more information about it.
[1] <http://www.topicmapslab.de/publications/mapmark>
[2] <https://code.google.com/p/mio/>
Best regards,
Lars
--
Semagia
<http://www.semagia.com>