Appendix A: Open Courses#
The idea of Open Courses is to take some of the principles of Open Research and adapt for the creation of online training material. Open Research is a set of principles to promote sharing research outcomes in a way that makes research reproducible and more transparent. There are at least four aspects that are considered:
Open Code (Open source licenses)
FAIR Data (Findable Accessible Interoperable Reusable)
Research Profile (e.g. ORCID)
Open Access of publications (Green, Gold, Hybrid, Diamond)
New ways of publishing potentially needs to consider all of the aforementioned points as the objective is to publish training material online with maximum reach. For transparency and accessibility the source code of the course should be open and any data used in the course should adhere to best practices. The authors of the course may also benefit from an open research profile which would give credibility to their content, transparency and increase the visibility to other work created by the same authors.
Open Code#
As indicated for the datasets, publishing the code of the training material online does not entitle the way in which the course material can be used. When the course is created it is automatically protected by copyright. The course source code material without a license can not be used. It is important then to understand what are the licenses available and to add them to the course material properly. The information provided in this roadmap is a personal overview and should not be taken as legal advise, additional information can be found at https://opensource.guide/legal/. Furthermore, an interactive guide on how to choose a license is available at https://choosealicense.com/.
One of the most common licenses for open-source is Creative Commons. It has
lots of variations depending on the flexibility that you want to provide to the
users. The Creative Common licenses are short named with acronyms, and
understanding them makes much easier their interpretation.
0 (No rights Reserved, Public Domain): There are no restrictions on how the code is used nor shared.
BY (By Attribution): Any use of the code needs to attribute the original work and author/s.
SA (Share-Alike): Any use of the code even if modified needs to keep the same license as the original work.
ND (No Derivatives): The work is shared as a whole and can not be modified.
NC (Non Commercial): The work can not be used for commercial uses.
The following are some common Creative Commons licenses:
CC-0: No Rights Reserved, allows the distribution without accreditation, it is commonly used to share tabular data or other databases from which knowledge could be derived.
CC-BY: Attribution, allows the use of the work even for commercial purposes but requires the attribution of the original form (e.g. with a citation). It is recommended for the widest dissemination of work.
CC-BY-SA: Attribution-ShareAlike, allows the use of the work even for commercial purposes but requires the attribution and the same type of license to any derivatives.
CC-BY-ND: Attribution-NoDerivatives, allows the use of the code as it is even for commercial uses, but does not allow the modification of the code. It also requires the accreditation of the original author.
CC-BY-NC: Attribution-NonCommercial), allows the use and modification of the code for non-commercial use, subject to accreditation of the author and does not require the same license on its derivatives.
CC-BY-NC-SA: Attribution-NonComercial-ShareAlike, allows the use and modification of the work for non commercial applications, requires accreditation of the original work and authors and the derived code needs to use the same license as the original.
CC-BY-NC-ND: Attribution-NonComercial-NoDerivatives, allows the use of the code without modifications for non commercial uses and requires accreditation of the original work and authors.
The Restrictive License Template: is a license developed by the Australian Government Open Access Licensing framework for material that contains personal or other confidential information. It can include multiple restrictions on its use like time limits, permissions or ethics required, or contractual arrangements). See more at https://library.unimelb.edu.au/Digital-Scholarship/restrictive-licence-template
Other common licenses for software are
MIT (license) allows the use of the Software free of charge, with no restrictions but under the condition that there is an accreditation of the original software and authors and that the permission notice is included in all the copies or substantial portions of the Software.
Apache (Apache License, Version 2.0) (license)
GPL (General Public License) guarantees the end users the four freedoms to run, study, share and modify the Software.
And the BSD license which includes several versions
BSD 0-clause (aka BSD Zero Clause License) allows the use, copy, modification and/or distribution of the Software for any purposes with or without fees.
BSD 2-clause license (aka “Simplified BSD License” or “FreeBSD License”) same as 0-clause but requires to retain the copyright notice, the list of conditions and a disclaimer in the source code and in the documentation or other materials provided if used in its binary form.
BSD 3-clause (aka “BSD License 2.0”, “Revised BSD License”, “New BSD License”, or “Modified BSD License”) same as 2-clause license but does not allow the endorsement or promotion of products in the name of the original copyright holders and contributors without specific prior written permission.
BSD 4-clause (aka original “BSD License”) same as 3-clause license but all advertising material that mentiones the use of the original sofware must display the following acknowledgement: This product includes software developed by the <copyright holder>.
FAIR Data#
The FAIR (Findable Accessible Interoperable Reusable) principles of sharing data were defined by a consortium of scientists and organisations, and published in the journal Scientific Data [WDA+16]. In STEM fields it is very common to share datasets with the students in order to better understand the training material. It is important to ensure that any data shared online in this manner is correctly licensed. The four principles that shape the acronym are:
Findable: Additional metadata is added in order to easily identify and find the data with a search engine. The metadata needs to contain clear information and the data requires a unique identifier.
Accessible: There is a specific protocol that can be followed to retrieve the metadata of the data of interest (even if the data is not available). The metadata is also understandable by humans and can be processed by machines. The data is stored in a trusted repository that ensures its accessibility for a particular period of time.
Interoperable: The metadata follows a formal structure that is commonly accepted by multiple parties. Ideally the metadata has a vocabulary that follows the FAIR principles and can form a knowledge representation.
Reusable: It is clearly specified how the data can be reused by others, including a license and any usage limitations.
Data format and storage#
Data should be stored in such a way that is easy to understand by humans, machine readable and accessible; including metadata human and computer readible. The data should be stored in open file formats facilitating the accessibility, and not requiring propertary applications to be opened. Should provide good documentation (e.g. a README plain text file), and be stored in a trusted dat repository for long term storage.
Some examples of data repositories are:
figshare allows to upload any file format and assigns a DOI identifier for citations.
Mendeley Data allows the storage of public or private data, keeps versioning and ensures long-term storage by Data Archiving & Networked Services2.
Zenodo is a general-purpose open-access repository that facilitates making the repository private and can automatically make it open once an associated paper is published. No restrictions on the file type and datasets up to 50 GB.
The Open Science Framework3 is an open-source research management and collaboration tool to facilitate the documentation of a full project lifecycle.
Data Licenses#
Publishing a dataset online does not entitle the ways in which the data can be used. When the data is generated it is automatically protected by copyright and without a license, users do not have permission to access, use and share under copyright and database laws. For example, in 2015, ~54% of the open data in open data catalogues across the European Union was not truly open as it was not licensed1.
There are multiple licenses to choose from, and as legal documents they may differ in some small details. In order to facilitate the understanding of the licenses and maximise the use of your material it is advisable to use well known licenses.
Some of the most common licenses are the Creative Commons. Similar principles shown previously in the open-source licenses section apply here. For example the Open Data Commons by Attribution License (ODC-BY v4.0) allows to copy, distribute and use a database, produce new works derived from it, modify, transform and build upon it subject to attribution to the original work. A similar license that forces any derived work to be shared under the same license is the Creative Commons by Attribution share-alike (CC-BY-SA v4.0). On the other hand, a less restrictive license for public domain is the CC0 which does not require attribution.
The licenses can also be created to fit particular purposes, or to facilitate its adoption in certain organisations. This is the case for some governments may create their own open data licenses to facilitate its adoption among governmental organisations. Some examples of governmental licenses are the UK Open Government License4, the French Government open License5 and the Singapore Open Data Licence6.
The European Union offers an online data licensing assistant7
The following video provides some quick guidelines on Data Sharing by the UK Reproducibility Network. The animation is shared under a CC BY license. Other UKRN primers are available at https://www,ukrn.org/primers.
Research Profile#
The authoring of any material should clearly and uniquely identify the identity of the contributors. In printing material it was common to use the name and surnames of the authors as an identification. However, in multiple occasions this is not enough for disambiguation. With more recent online tools it is possible to disambiguate authors by providing additional information and creating unique identifiers for each of them. Webpages like ORCID8 (Open Researcher and Contributor ID) provide unique identifiers for researchers and connect them with their contributions. Other identifiers are the Scopus Author Profile9 by Elsevier, ResearcherID10 from Clarivate, the Digital Author Identifier (DAI)11 by the Dutch research system, among others.
Open Access#
We provide a short introduction to Open Access for completeness of the Open Research principles, which is mainly directed for research publications. However, we believe this principle may be considered if the training material is reused for a research publication. Open Access is a broad international movement that specifies a set of principles to make research free of access and online, defined as “digital, online, free of charge, and free of most copyright and licensing restrictions.”. One of the main objectives is to publish research work under the CC-BY license. If the material for a course is later reused to publish a book or some other type of publication it may be necessary to know what are the possible licensing restrictions.
Fig. 13 Open Access logo#
There are multiple types of Open Access models that are used by journals. However, the most common ones are gold, green and hybrid OA journals, while some journals are hybrid. In all the cases the article is free to read but they have the following differences.
Gold OA: The article is free to read, but the authors need to pay the publisher, requiring an external funding. The ‘Article Processing Charge’ costs an average of £2,000 and can reach £10,000.
Green OA: The author accepted manuscript can be hosted in some repository (e.g. Pure is a repository of scholarly works for the Universtiy of Bristol). The publisher retains the final version in their website. The article has a CC-BY license and there is no embargo on its use. Additional considerations may be required by the authors, for example the inclusion of a statement in the publishers submission, like in the case of the University of Bristol recomendation:
"For the purpose of open access, the author(s) has applied a Creative Commons
Attribution (CC BY) licence to any Author Accepted Manuscript version arising
from this submission."
Diamond OA: These are crowd-funded by libraries and scholarly organisations that pay for the processing charges. Then in the same manner as the Green OA, the accepted manuscript can be hosted in a repository, while the final copy is available in the publisher’s website.
More about Open Research#
Additional sources of information about open research can be found in the following links: