[Labs-l] help/suggestions on how to set up a triplestore to query a database of etymological relationships

[Ester Pantaleo]

Hi all,

I am developing an IEG project to visualize the etymology of words
using data extracted from the English Wiktionary and I need to set up
a DBMS (triplestore) to query ~3Gb of triples.

Tlhe inks to the project proposal and to the demo are:

https://meta.wikimedia.org/wiki/Grants:IEG/A_graphical_and_interactive_etymology_dictionary_based_on_Wiktionary
http://tools.wmflabs.org/etytree/

So far I have used a Virtuoso server http://kaiko.getalp.org/sparql
which is not managed by me and it has become difficult now to update
it.

I would like to set up a triplestore for my data (~3 Gb but the size
can be reduced after filtering some data I don't really need). I was
thinking of using Blazegraph - some people at wikimedia recommended
using it - or Virtuoso (setup for Virtuoso is described here:
http://kaiko.getalp.org/about-dbnary/setting-up-a-virtuoso-opensource-server-to-mirror-dbnary-data/
).

Does any of you have recommendations on which DBMS to use? Anyone
willing to help setting this up? I opened a request on Phabricator:
https://phabricator.wikimedia.org/T151762

Any feedback/help very welcome :)

Ester