Hi Raffaele,
On 10/03/14 13:38, Raffaele Palmieri wrote:
From my understanding, there are to way to understand how to "integrate
OpenRefine with Marmotta":
1. Allow to push data to Marmotta from OpenRefine, basically port what we
did in LMF a couple of years ago: http://goo.gl/awQh5z
I think that's the first step, because in the meantime Google Refine
became Open Refine, so it is necessary re-check that mode of integration.
Moreover, if I remember correctly, Google Refine expected that extensions
are built with Ant, so a work about extension configuration needs to be
done to eventually migrate to Maven.
The re-branding has not affected the core functionalities: besides some
package renaming, the old extension should work completely fine in the
latest versions. That's why I don see just that as a target project for
GSoC, unless you include something else, like MARMOTTA-202, read below.
Extensions can internally use Maven, but does not really affect the
runtime. Anyway for quite a long I've trying to convince the community
about such switch, but they look more worried about touching something
that works than the advantages. Not blocking anyway.
2. My original motivation for MARMOTTA-202 was to offer a headless
integration of the OpenRefine engines for transforming data to RDF.
What do you mean for headless integration? A similar work to GRefine RDF
Extension?
Maybe I did not describe properly MARMOTTA-202... Well, in the end the
idea is to allow Marmotta to import CSV files together the script you
could generate from OpenRefine. That would allow you to import in batch
several files sharing share structure.
The second one was more relevant for me last year, because in the meantime
we (Salzburg Research) cooperate in some research activities (
http://www.fusepool.eu/p3) that have something similar in the pipeline,
most-likely open source.
It would be nice to see how you mean integration into the project Fusepool
and work towards a common goal.
And I'll try, for sure, but GSoC does not really fit with the timing
there. But, as I said, we could experiment in GSoC and later converge.
Raffaele, could you be interested on mentoring it?
I could try, if we clarify all doubts about expected job, it would be
easier.
At high level, MARMOTTA-202 could be summarized: "Import RDF data into
Marmotta from a set of CSV files with a shared structure taking a
template exported from OpenRefine, including the RDF mappings". Then is
clear some internal parts of OpenRefine would need to be used, but
further details would come during the proposal preparation and the
mentoring; right now I did not go further to know the actual features
and limitations of such idea.
Cheers,
--
Sergio Fernández
Senior Researcher
Knowledge and Media Technologies
Salzburg Research Forschungsgesellschaft mbH
Jakob-Haringer-Straße 5/3 | 5020 Salzburg, Austria
T: +43 662 2288 318 | M: +43 660 2747 925
[email protected]
http://www.salzburgresearch.at
