Data Licensing

April 27, 2011

Data Licensing has slowly been growing as a topic within the MeCAT project.  From my perspective, adopting an existing license framework is the most attractive options as:

  • Writing something from scratch will be expensive and time consuming.
  • The existing framework should be well tested, thus more likely to meet expectations of data protection.
  • An existing framework should be well known, i.e. users of the data will (hopefully) recognise the license and know what it means without further (time consuming) research of the license (thus promoting re-use of the data, which otherwise might simply be too much trouble).

ANDS (our project funder) has a few pages on licensing, starting at Data Re-use and Licensing Frameworks.  Since they are paying the bills, it is an obvious place to start and a default first choice.  My understanding of where they are currently heading is GILF (AusGOAL), which is basically the Creative Commons licenses, a Restrictive License framework and a framework on how to apply it.

The UK Digital Curation Centre (DCC) How to License Research Data provides a good summary of the various license frameworks.  Creative Commons certainly seems to be the leader based on my criteria above.

Creative Commons has been used for data and databases over the years, but data hasn’t been a focus for the organisation.  That appears to be changing, based on the Creative Commons blog entry CC and data[bases]: huge in 2011, what you can do:  CC licenses are being encouraged and adapted for scientific data

… with the important caveat that CC 3.0 license conditions do not extend to “protect” a database that is otherwise uncopyrightable.

As others have pointed out (link needed), GILF is aimed at Government departments, and so while the basic framework seems appropriate, some of the GILF Policy wording is potentially problematic in its reference to minimum requirements for agencies, employee requirements, etc.  ANDS is working with GILF to improve the wording.

So where does that leave MeCAT?  In the short term, I think we’ll use the CC 0 – 6 licenses as a starting point and review the decision as ANDS updates its position.  The Restrictive License would be useful as a starting point for researchers wanting to share data (only) with other researchers, however it isn’t an urgent consideration for us at the moment.

Attribution wording, liability, etc. are all still to be worked out.