Presentation given by Gareth Knight on topic of data licensing at London School of Hygiene & Tropical Medicine on May 26th, 2015

1. INTRODUCTION TO
DATA LICENCES
Text is licensed under a Creative
Commons Attribution 4.0 International
License. Other licences apply to
images (see slide 19)
Gareth Knight
gareth.knight@lshtm.ac.uk
26th May 2015
Disclaimer:
I am Not a Lawyer. This is not legal advice

2. Overview
• Reasons to apply a data licence
• Database rights
• Expectations for data licences (LSHTM, funder, journal)
• Licensing your research data (CC, ODC, DTAs)
• Working with 3rd party data
• Conclusion

3. Data Licence
• Data licence motivated by need to
share data
• Establish conditions under which
data may be accessed, used & cited
• Many rights apply to data
– Copyright
– Moral
– Database
• Rights issues vary between countries
• Ensure that your project has clarified
rights issues before sharing

4. Database Rights
• Does not require 'creative' aspect
defined by copyright to be protected
• ‘Copyright and Rights in Databases
Regulations’ 1997 applies it to:
– DBs created after 27 March 1996
– Created by EEA nationals, EEA
residents & businesses with central
operations in EEA
• Last for 15 years from completion or
publication (whichever is longest)
• Any substantial update causes the 15
year period to begin anew
Recognise investment made into compilation of database

5. LSHTM Expectations
• Rights information should be:
– Clear and unambiguous
– Documented at an appropriate level of
granularity
• Recommend a non-exclusive licence
that allows many people to access and
use data
• Should not assign exclusive rights to 3rd
party, unless it is a condition imposed
by contractual or other obligations.
“Rights assigned to research data should not unnecessarily
restrict its management, sharing, or use”
RDM Policy, Principle 2
LSHTM RDM Policy
http://www.lshtm.ac.uk/research/researchdataman/rdm_policy.html#principle02
LSHTM RDM Policy
http://www.lshtm.ac.uk/research/researchdataman/rdm_policy.html#principle02

6. • Many state data should be:
"made openly available and accessible with
as few restrictions as possible"
• Encourage FAIR principles - Findable,
Accessible, Interoperable & Re-usable
• Expect consortium agreements to enable
data sharing
• However, recognise legal, ethical &
commercial constraints may limit access
and use
• Encourage licences that encourage uptake
in some form (Open, controlled,
commercial)
Funder Expectations

7. Journal Expectations
• Growing number of journals that:
– Specialise in open data
– Expect data to be made available
• Expect openness by default, unless
good reason can be provided
– Creative Commons Attribution
(CC-BY)
– Anonymous download from a
public repository
• Recognise open sharing not always
feasible for health data
– Controlled access
– Data sharing agreements

8. Licensing your research data
1. Identify rights holders
– LSHTM
– Collaborators
– Participants
– Data providers & others
2. Determine terms to be applied
– Policy, Consortium agreement,
consent forms
3. Review licence models
– Creative, Commons
– Open Government Licence
– Data Transfer Agreement
4. Select appropriate licence

9. Creative Commons
ATTRIBUTION
NON-COMMERCIAL
NO DERIVATIVES
SHARE ALIKE
+ : Permits open reuse
- : ‘Attribute stacking’ – distant authors
- : Unable to control how data is used
+ : Permits academic & other NC use
- : Cannot be used by commercial
collaborators
+ : Allows analysis
- : No cleaned, remixed, or other
derivatives allowed
+ : Prevents future users re-licensing
derivatives under a restrictive licence
- : Difficult to combine with other licences

10. Open Data Commons
Share data
Create works
Adapt & modify
ODC Open Database License (ODbL)
http://opendatacommons.org/licenses/odbl/
Attribution, Share alike & keep open
ODC Attribution
http://opendatacommons.org/licenses/by/
Attribution
ODC Public Domain Dedication
http://opendatacommons.org/licenses/pddl/summary/
None

11. Data Transfer Agreement
Tailored licence form that defines
conditions such as:
• Use for specific purpose
• Handling of participant information
(no attempt to re-identify)
• Storage in secure environment
• Retention period (e.g. 1 yr)
• Administration charges
Signed agreement that specifies storage & use conditions
Contact the Knowledge Transfer Manager (deborah.carter@lshtm.ac.uk) for guidanceContact the Knowledge Transfer Manager (deborah.carter@lshtm.ac.uk) for guidance

12. Working with 3rd party data
Myth:
• Data has no licence: It can be analysed for any purpose
• Data is free to download: It can be re-published in any form
Review dataset description for licence information:
1. What purposes are you allowed to use data for?
2. Are there any expectations?
– e.g. citation, costs
3. Are there any limitations?
• If there's no licence, contact the creator/publisher.
Large amount of data available, but can you use it?

13. Dealing with multiple datasets
Potential for licence conflict when
working with multiple datasets
Licence compatibility definitions:
• Exact match: Both licences use terms
that have the same purpose, meaning
and effect
• One-way: Data made available under a
permissive license can be combined with
data that has a more restrictive licence
Does source licence allow mashups to be shared?
http://creativecommons.org/compatiblelicenses
http://opendefinition.org/licenses/
http://opensource.com/law/11/9/mpl-20-copyleft-and-license-compatibility

14. Creative Commons:
No Derivatives
Instructions needed to explains how dataset
may be merged for validation purposes
Based upon scenarios by Leigh Dodds "Oil and Water: When Data Licences Don't Mix“
http://www.slideshare.net/ldodds/oil-and-water-when-data-licences-dont-mix

15. Creative Commons:
Share alike Conflict
Each update can be published separately & instructions
provided on how to merge for validation
Based upon scenarios by Leigh Dodds "Oil and Water: When Data Licences Don't Mix“
http://www.slideshare.net/ldodds/oil-and-water-when-data-licences-dont-mix

16. Creative Commons:
Mixing open licences
If licences are compatible, least open licence wins
Based upon scenarios by Leigh Dodds "Oil and Water: When Data Licences Don't Mix“
http://www.slideshare.net/ldodds/oil-and-water-when-data-licences-dont-mix

17. Conclusion
• Licence models are a key component of data sharing
• Researchers must consider data licences when
– Reusing existing data
– Creating own data
(particularly when there is a need make data available)
• Rights information should be clear, unambiguous and documented at
appropriate granularity
• Recommend use of a non-exclusive licence that allows many people
to access and use data
• Should not assign exclusive rights to 3rd party, unless it is a condition
imposed by contractual or other obligations.

18. Resources
• LSHTM. Choose a Data Licence
http://www.lshtm.ac.uk/research/researchdataman/share/choose_licence.html
• Digital Curation Centre: How to License Research Data
http://www.dcc.ac.uk/resources/how-guides/license-research-data
• Korn & Oppenheim (2011). Licensing Open Data: A Practical Guide
http://discovery.ac.uk/files/pdf/Licensing_Open_Data_A_Practical_Guide.pdf
• ODI. Publisher’s Guide to Open Data Licensing
https://theodi.org/guides/publishers-guide-open-data-licensing

19. Images
Slide 3: “That’s Right” by Kaytee Riek (CC BY-NC-SA 2.0)
https://www.flickr.com/photos/riekhavoc/4813140176/
Slide 4: “WordPress 2.7 Database Schema” by Rafael Poveda (CC BY-NC-SA 2.0)
https://www.flickr.com/photos/bioxid/3640432505/
Slide 8: “Reuse” by Steev Hise (CC BY-NC-SA 2.0)
https://www.flickr.com/photos/steev/39393264/
Slide 11: “Permit Holders Only A” by Gregory Wake (CC BY-NC-SA 2.0)
https://www.flickr.com/photos/gregwake/2301264039
Slide 12: “Sharing” by ryancr (CC BY-NC 2.0)
https://www.flickr.com/photos/ryanr/142455033/
Slide 13: “Death By…” by nataliej (CC BY-NC 2.0)
https://www.flickr.com/photos/nataliejohnson/2296566285/