[From Alan Ball]
The Digital Curation Centre (DCC) has published the second in its series of
How-to Guides: 'How to License Research Data' by Alex Ball of the DCC, in
association with JISC Legal. The guide explains why licensing data is
important, what licensing options are available to researchers, and how to go
about attaching a licence to a dataset.
The DCC's How-to Guides offer practical introductions for those who need more
than the high-level basic awareness given in DCC briefing papers, but less than
the in-depth coverage given in the Curation Reference Manual. This guide is
aimed at principal investigators, researchers, and those who provide access to
research data through a data centre, repository or archive.
'How to License Research Data' is available for online reading or download
This is a valuable medium-level overview of the different legal aspects of publishing data. Data, of course, is now found in many disciplines , not just science. At the very fine-grained level, data is extraordinarily complex, but at a high level it can be very easy.
If you are a scientist and haven't thought about data licences, then consider the value of making your data available to others. That's anathema to many traditional scientists – and an attitude that will survive for some time. There are moral, ethical, political, social and utilitarian reasons why you should consider making your data Open. [There are cases where you cannot open data – human privacy, breeding grounds of rare species, etc. And often the decision involves other people. But at least consider it.]
I am not a fan of licences. They are complex, legal algorithms do not map onto any formal system of mathematics, and they are subject to wide variation by country, date and general fuzz. That's why wherever possible you should adopt the Panton Principles (www.pantonprinciples.org/
) and formally dedicate your data to the Public Domain (PDDL or CC0). The complexity of combining even two licences is far greater than analysing a terabyte of multivariate data for patterns. Multiple licences make data recombination very hard.
AND NEVER USE NON-COMMERCIAL LICENCES.
The only people this hinders and hurts are people like you.