What license should I choose when publishing data in SND (Swedish National Data Service)?

In short

• The general recommendation is that SLU datasets published openly in SND are released under CC0
  (Creative Commons Public Domain Dedication, which is not a license but a waiver/mark).
   
• For datasets with restricted access, no license or waiver should be applied since they should not be disseminated.
   
• If you have used existing data collected by others, or if data stem from a project where the use or publication of the data is regulated by an agreement, you need to check conditions before considering publishing them.
   

Recommendations on this page are developed in collaboration between Data Management Support and SLU Legal Affairs, based on internal and external guidelines and policies. Please read more about CC0 and licenses as well as guidelines and policies below and follow the links for further information. Contact DMS if you have questions.
 

Background

Providing clear conditions for the reuse of published datasets is important for data to be reusable according to the FAIR guiding principles for scientific data management. It is the policy of SLU as well as the Swedish government, EU and many funders and publishers that research data be published as openly as possible and with no more restrictions for reuse than necessary.

• FAIR principles
• SLU data management policy

About CC0

At SLU, the general recommendation is that SLU-generated datasets that are to be published openly in SND are released under the CC0 waiver (Creative Commons Public Domain Dedication). This recommendation is based on internal and external policies and guidelines stating that data should be shared with as few restrictions as possible. SLU and SND endorse the recommendations made by Digg (the Swedish Agency for Digital Governance) that data from the public sector should be released under CC0 or PDM (Public Domain Mark) when no copyright or related rights are applicable. 

• Digg recommendations on open licences (Swedish only)
• About CC0 (Creative Commons)
• About CC0 for data (Creative Commons)

Another reason for why CC0 is recommended when publishing data is the fact that it is difficult to ascertain whether or not datasets are subject to copyright. Contrary to research articles that are literary works, information and data as such cannot be copyrighted nor owned (see SLU’s policy on intellectual property).

Applying CC0 is a way to remove any doubt about whether and how others, in any jurisdiction, can reuse the published data material. CC0 is not a license but a mark that makes it clear that others can share and adapt (distribute, remix, adapt, and build upon) the published material without asking for permission, thereby facilitating the reuse of data as much as possible. CC0 is widely used and recognised in the scientific community and is applied by many research data repositories such as Dryad (where CC0 is mandatory), Figshare and Harvard Dataverse (where CC0 is the default). CC0 is also commonly applied among Swedish authorities such as Naturvårdsverket (Swedish Environmental Protection Agency), Jordbruksverket (Swedish Board of Agriculture) and Havs- och vattenmyndigheten (Swedish Agency for Marine and Water Management).

Although applying CC0 does not legally require users to cite the source, attribution is still an important part of ethical norms and good research practice. Before downloading data from SND, users have to confirm that they have read guidelines about giving credit to data creators and citing datasets.  

• SND guidelines for reusing and citing data

About CC BY and related licenses

Licenses can be applied to material that is subject to known copyright or similar rights. Data may be covered by such rights if they consist of creative works such as elaborate texts or images, or to data that have been compiled into a database meeting the criteria of either creative originality or of a substantial investment in the obtaining, verification or presentation of its contents (not in the creation of the contents or generation of the data as such). Among the CC licenses, CC BY (Attribution) is the general recommendation. However, the definition of a database in this sense, as well as the required level of originality and investment, are rather unclear and may vary depending on jurisdiction, which is one of the reasons why the CC0 waiver is the main recommendation for datasets. 

Please note that licenses can only be applied by the one holding the rights. If you consider applying licenses that state more restrictions and conditions for reuse than CC BY, such as CC BY-ND (Attribution-No derivatives) or CC BY-NC (Attribution-Non commercial), you need to be able to justify these restrictions from a legal or commercial point of view.

• More information about the CC licenses (Creative Commons)
• Considerations for licensors and licensees (Creative Commons)

Note that licenses do not have any legal effect if no copyright is applicable, and that they regulate the reuse and dissemination of the published material, not of the data/pieces of information in and of themselves.

CC licenses may also be applicable to documentation files if they are subject to copyright (e.g. journal articles) - see "Multiple material" below.
 

Code and software

For code and software, licenses other than Creative Commons are more suitable. Licenses meeting open source code standards are recommended. Linked below is a tool to help you choose the right license.

• Choose a license for open source code
   

Multiple material with separate terms and conditions for reuse

Note that what you choose in the "License" field will apply to the whole of the data submission, including all data files and documentation files, unless otherwise specified. If, for instance, uploaded documentation and/or code should be associated with different terms than the data files, you can enter "Other" in the "License" field and describe the terms for each file in a readme (readme.txt) or a license file (license.txt). A clarifying comment in the field for "Copyright" may be useful.

• Example of license information in a license.txt file
• Example of license information in a readme.txt file

Data with access restrictions

If the data material is to be placed under access restrictions for legal, ethical, security or commercial reasons, it should not be disseminated (though it may be handed out on request depending on transparency and secrecy legislation). It is therefore recommended to leave the "License" field empty.