Announce: RDF::Generator::Void 0.12

[Kjetil Kjernsmo]

All,

I just came to think of, I don't think have yet announced my VoID generator
Perl module to the void-discussion list, and I should. Basically, this module
takes a number of arguments to allow you to set certain things like endpoint
and stuff manually, and it does some heuristics, like that 1% of prefixes of the
URIs used in properties are likely to be vocabularies (can be overridden).
Most importantly, however, it does a single pass of the triplestore to compile
propertyPartition and classPartition statistics. This is the statistics needed
for federation query optimization as used in the SPLENDID paper by Görlitz and
Staab (for those who know the scientific literature).

It is easily integrated in a web server (like it's done with RDF::LinkedData
which runs on http://data.lenka.no/ where it runs in real time), but the new
release also adds an optional command line tool.

So, for the recent release, I have uploaded file

RDF-Generator-Void-0.12.tar.gz

which has entered CPAN as

file: $CPAN/authors/id/K/KJ/KJETILK/RDF-Generator-Void-0.12.tar.gz
size: 102017 bytes
md5: da85b71eb98a5aec263aa8f715d954e6

Excerpt from the changelog:
0.12 2013-07-07 # Some provenance and command line tool.

- (Addition) Author tests changes.
- (Addition) Command-line usage through perlrdf.
- (Addition) Use provenance ontology to assert that this module generated
the description.

For those who are not familiar with CPAN, you can find the module here:
https://metacpan.org/release/RDF-Generator-Void

It is also available in Debian and Ubuntu. The version in Debian stable and
Ubuntu quantal is 0.04, while this release is in Debian testing and unstable
and Ubuntu Saucy.

Perl is usually not very fast, but it works well enough for the present
purpose. I have just upgraded my box with the intention of running it on the
DBPedia ontology data, so we'll see how well it performs on larger datasets
soon :-)

Share and enjoy,

Kjetil