Other threads on this site (e.g. #3, #6, #8) discuss how to model MELODIES data and make it discoverable and linked. But where should we host the actual data? In some cases (e.g. WP3) there may be a natural home for the data (e.g. a national data centre), but this may not be the case for others. Options may include:
The MELODIES website - but this will probably only be a temporary thing.
datahub.io - [this is actually down at the moment!] - TELEIOS data is already published here.
zenodo.org - data gets a DOI. Seems to link somehow with EU funding through OpenAIRE.
FigShare - probably too "informal" for our main outputs.
A CKAN repository somewhere (where?)
In any case, I guess that we want the flexibility to describe the data how we want, using the ontologies etc. that we prefer. Perhaps one approach would be to publish metadata on the MELODIES website (perhaps as DCAT, VoID, schema.org or some vocabularies like that) and host the data elsewhere (e.g. zenodo), using the DOI as an identifier to link the two.
Other threads on this site (e.g. #3, #6, #8) discuss how to model MELODIES data and make it discoverable and linked. But where should we host the actual data? In some cases (e.g. WP3) there may be a natural home for the data (e.g. a national data centre), but this may not be the case for others. Options may include:
In any case, I guess that we want the flexibility to describe the data how we want, using the ontologies etc. that we prefer. Perhaps one approach would be to publish metadata on the MELODIES website (perhaps as DCAT, VoID, schema.org or some vocabularies like that) and host the data elsewhere (e.g. zenodo), using the DOI as an identifier to link the two.