The most common means for the owners of rights in datasets to grant permission is through a licence. A licence is a legal contract that specifies standard restrictions and permissions for a work, which can then be used by anyone the licence is granted to, although the ownership rights are not transferred. Again, keep in mind that licences usually regulate only the copyright and copyright related aspects. Personal data protection cannot be addressed through licences.

Licences might have been applied to data in a number of ways: Embedded in the metadata for the data; Communicated by watermarks or notices within the data; Specified on the landing page for the dataset; Specified on a repository website; Detailed in a ReadMe file released with the dataset.

Using data under a licence can become more complex when that data may be combined with data from other sources. The result may be a dataset with different licensing restrictions. The ability of licences to interact with other licences is called interoperability. Not all licences can accommodate different layers of protected work. Creative Commons licences are better at this than some other licences, and bespoke licences can present particular interoperability issues.

When choosing a licence for protected research data, it is necessary to consider how this might interact across the whole research data (including non-protected elements) and how it might interact with future derivative works produced from re-use or re-purposing.

OpenMinTeD has a useful matrix presenting the compatibility of different licences. Using two or more licences may require stacking of attribution of rights in the licensed work.

Not necessarily, but it depends on how you use the SA licensed content. A ‘Share Alike’ CC licence applies only to the content licensed as SA that you have used. It does not require you to also make your work available under a SA licence, so long as you have not combined the independent works into one new work (known as a ‘derivative’ work).

If the dataset is not protected by copyright or related rights (see above) yes. If they are protected, then in the absence of a licence the owner(s) of rights must be approached directly for permission. There are a limited number of exceptions that permit use without a licence or specific permission. Examples of exceptions for copyright protected works include: Research & Private study; Quotation, Criticism and review; Disability access; Public administration; Text and data mining; News reporting; Parody & Pastiche. As said, exceptions and limitations to copyright are not harmonised at the EU level, so you should verify which ones are available to you. Also, keep in mind that these exemptions are granted in limited circumstances and subject to certain conditions. As usual, specific advice is recommended before relying on an exception.

Where a dataset has protected elements, intellectual property laws may permit the owner of the rights to take steps to preserve their exclusive rights to use, copy and make derivative works from the data. Ultimately, the rightsholder could use the courts to enforce these rights, by asking a judge to make an order to stop use of the dataset, to destroy copies or derivative works, or to pay compensation to the rightsholder. The time and cost of any legal action, and the risk of an adverse judgment, may outweigh the benefits of using the dataset.