20200504_Data, Data Ownership and Open Science

Data, Data Ownership and Open
Science
OpenAIRE Legal Policy Webinars
4 May 2020
Dr. Thomas Margoni
Senior Lecturer in Intellectual Property and Internet Law
Co-director, CREATe
Coordinator, IP LLM Programme
School of Law – CREATe
University of Glasgow
www.create.ac.uk
thomas.margoni@glasgow.ac.uk

Example: OpenMinTeD
Copyright theory (and sometimes copyright law) says: no
E.g.: ideas, procedures, methods of operation, mathematical concepts, etc are
not protected by copyright law, only original expressions which constitute
intellectual creations (e.g. Arts. 2 WCT, 9(2) TRIPs, Art. 2 Berne and most
legal traditions requiring originality). Factual information and data as such fail
to qualify for copyright protection.
Other areas of law may say yes, but usually in specifically identified
situations, or with limited remedies e.g.:
1) Trade secrets, confidentiality (only if secret and limited remedies)
2) Contracts (privity, enforceability and remedies)
3) Data protection (only qualifying data and scope is protection)
4) PSI (reuse by default)
5) …
Is data owned?
thomas.margoni@glasgow.ac.ukthomas.margoni@glasgow.ac.uk

Example: OpenMinTeD
Databases: Are protected by copyright if the selection or arrangement is
original. However, what is protected in this case is the selection or
arrangement (e.g. the structure of the database), not the contained data.
● This means that if the original database is composed of copyright
protected elements (e.g. a DB of journal articles, films, songs, etc) the
content of the DB is protected NOT because is data, but because they
are works of authorship in their own right.
● If the original database is composed of non protected information (e.g.
temperature measurements, numbers or other factual information), the
copyright protecting the structure does NOT extend to the underlying
data, which remains free to reuse from a copyright point of view (but you
should always verify this is not personal or confidential data, etc.).
Is data owned?

Example: OpenMinTeD
Databases: In the EU, non original databases are also protected (heard of the
SGDR?). In this case, a database whose making required a substantial
investment in obtaining, verifying or presenting (NOT creating) the data is
protected by Sui Generis Database Right (SGDR, a “lighter” form of
copyright), therefore effectively protecting obtained (but not created) data.
Copyright and SGDR are cumulable, so an original database (structure)
whose making required a substantial investment may enjoy a double form of
protection. If the “data” are in fact works (a database of journal articles) there
are 3 layers of rights protecting that database. All of them have to be properly
authorised if you want to reuse that database (e.g. this is way it is important to
use the right licence, e.g. CC BY 4.0).
Is data owned?

Example: OpenMinTeD
What about the “data” contained in a work (not in a database)? Can you
extract statistical information about the English language from, e.g. a Harry
Potter novel?
In theory yes (the right to read is the right to mine), but in practice no (you
need a TDM exception).
If we take a closer look, EU copyright law (doesn’t say anything explicit
but) often means: YES.
Interplay between Arts. 2 and 5 InfoSoc Directive requires to obtain an
authorisation for data capturing/extraction + SGDR.
Is data owned?

Example: OpenMinTeD
Modern data analytics (e.g. TDM, machine learning etc.) normally extracts principles,
facts, data, correlations, etc, which copyright theory stipulates that are not protected,
thus the extraction of those unprotected elements from protected works should not
need an exception (non consumptive uses).
Main obstacle of EU copyright framework is that it harmonised rights broadly
(reproduction, redistribution, communication to the public, etc), but did not do the same
with exceptions (exhaustive but not mandatory list, narrow interpretation, etc). There is
no flexible standard in the EU such as there are in many other countries around the
world.
Result: Often data contained in works are de facto or de jure (for non original
databases) protected (needs authorisation) in the EU but may not abroad.
Is data owned?

Example: OpenMinTeD
By excluding protection of ideas, principles, factual information, non original
expressions the law’s goal is avoid the creation of monopolies over the
information needed by everyone to think, communication and create new
knowledge and to avoid the distortion of scientific freedoms and fundamental
rights that it may cause.
By excluding data created in databases the law’s goal is to avoid as much as
possible so-called “single source databases” due to their anti-competitive and
monopolistic nature.
Ideas and data as such are the basic bricks of human knowledge and should
not be owned but held in the public domain for everyone to have access to it,
but also to verify, reuse and replicate.
By affording protection to the obtaining of data structured in qualifying
databases against substantial extractions a limited reward for the investment
is given to the maker under EU law.
Is data owned
and why does it matter for OS?

Example: OpenMinTeD
1) Text and Data mining: computational analysis of anything recorded in the
work (sec. 29A CDPA) or any automated analytical technique aiming to
analyse text and data in digital form in order to generate information such as
patterns, trends and correlations (Art. 3 CDSM);
2) Scope: exception to the right of reproduction (both);
3) Beneficiaries: Non commercial research (29A CDPA), research
organisations with lawful access for research purposes (Art. 3 CDSM), anyone
for any purpose but can be opted-out (Art. 4 CDSM).
4) Relationship to contracts: Cannot be limited by contract (except for Art. 4
CDSM).
5) Relationship to technology: Can be limited by technological measures
(integrity measures and TPM)
Example: TDM exceptions

Example: OpenMinTeD
OpenAIRE has 3 published guides for researchers at the moment:
● How do I know if my research data is protected?
https://www.openaire.eu/how-do-i-know-if-my-research-data-is-protected
● How do I license my research data?
https://www.openaire.eu/how-do-i-license-my-research-data
● Can I reuse someone else’s research data?
https://www.openaire.eu/can-i-reuse-someone-else-research-data
And a companion guide to help address open science issue for repositories:
● Making your repository Open
https://www.openaire.eu/making-your-repository-open
CREATe Open Science resource:
● https://www.create.ac.uk/open-science/
Guides

Example: OpenMinTeD
● Open Access: Everything under CC BY 4.0/CC0 (or similar) can be copied,
reused, redistributed and mined (e.g. Wikipedia, www.copyrightuser.org) for any
purpose
● OpenCovidPledge: “to make our intellectual property available free of charge
for use in ending the COVID-19 pandemic”: www.opencovidpledge.org with
companies such as Amazon, Facebook, Hewlett Packard Enterprise, IBM,
Microsoft declaring adoption of either or both licenses.
● Wellcome statement on ‘Sharing research data and findings relevant to the
novel coronavirus (COVID-19) outbreak’:
https://wellcome.ac.uk/press-release/sharing-research-data-and-findings-relevan
t-novel-coronavirus-covid-19-outbreak
● Joint Comment to WIPO on Copyright and Artificial Intelligence,
http://infojustice.org/archives/42009#members
Recent initiatives

Example: OpenMinTeD
● Q: “How can we guarantee the reuse of data.”
A: Short answer: CC0 (long answer is explained in the Guides and depends on if
it’s just raw data, an original database, a non original database, etc).
● Q: “… the Open Data directive and IPR may conflict. For instance, when
data published by a government body are re-used by research consortium
consisting of a public-private partnership, I could imagine that the
consortium would want to restrict access to the (processed) data, claiming
intellectual property rights, while the original data ought to be open,
based on the Open Data directive…”.
A: Depends on a number of factors. In the simplest sceario Open Data requires
reusability. If there are no rights covering the data, then they can be reused in
any way (including within a proprietary model). If data are protected then usually
they come with a licence (CC or Open Government) and that licence must be
followed.
Some questions (and tentative
answers...)

Example: OpenMinTeD
● “Q: How and under which 'umbrella' can data legally be protected, if so
required? Copyright, Database right, ownership right, protection of trade
secrets?”
A: Data as such cannot be protected by copyright or SGDR. Data structured in a
database may be protected by SGDR if there is substantial investment in
obtaining, verifying or presenting (not creating) data. Trade secret may protect
data but under certain conditions. Contracts as well.
● Q: “How can information be retrieved without being considered
plagiarism?”
A: Ideas and facts are not protected. Original expressions are. Do not copy
original expressions.
● Q: “İs it legal to summarise my thesis as an article that may publish in the
international journal?”
A: If it’s your thesis and you did not transfer your copyright (e.g. to a publisher)
you can do whatever you want (from a copyright point of view).
answers...)

Example: OpenMinTeD
● Q: “Im interested in how to make researchers understand that is important
to protect results (IPR) before publishing the results. Many of them are
extremely focused in publishing the results and eventually have some
problems when trying to protect them (via patent for example). Thanks in
advance”
A: In general Universities have (or should have) TTOs which advise on these
matters. In principle, if you disclose your invention before obtaining a patent, this
disqualifies you from obtaining a patent later on. This however only applies to
certain field (humanities are less affected by patents than bio-med). Also, it
depends on the overall strategy of the University. Patents are an important tool in
a IP portfolio, but should not be over-used, the risk is to lock in important
knowledge created with (very often) public funds.
It should be kept in mind that differently from Patents, Copyright (articles,
software, database, books, images, films, etc) and SGDR (database) work
automatically, so authors are automatically protected (no need to register). Once
you are protected publish with the right Open Access licence (e.g. CC0 or CC BY
4.0) to get the best mix of visibility, re-usability and also protection (e.g. the BY
requires attribution).
answers...)

Open Access Week 2019
Thank you!
Dr. Thomas Margoni
Senior Lecturer in Intellectual Property and Internet Law
Co-director, CREATe
Director, IP LLM Programme
School of Law – CREATe
University of Glasgow
www.create.ac.uk

20200504_Data, Data Ownership and Open Science

More Related Content

What's hot

Similar to 20200504_Data, Data Ownership and Open Science

More from OpenAIRE

Recently uploaded

20200504_Data, Data Ownership and Open Science