In August 2022, a new Swedish law governing accessibility to public sector data came into force (Lag (2022:818) om den offentliga sektorns tillgängliggörande av data). The law has provisions that apply to public bodies making research data accessible. The law implements the EU Directive on Open Data and the Re-use of Public Sector Information. The the open data law replaces a previous law (Lagen (2010:566) om vidareutnyttjande av handling från den offentliga förvaltningen), known as the PSI law.
Purpose of the legislation
The law aims to promote the public sector making data accessible so that it can be used for other purposes, in particular in the form of open data, on condition that requirements for information security and protection of personal data can be maintained, and that Sweden’s security is not compromised.
Actors that will be applying the legislation
The law is the starting point for public actors. For higher education institutions (HEIs) whose principal is national government, it will only apply for research data. The limitation to research data also applies for other public agencies whose activities include carrying out or funding research, or who make research data directly accessible. For such public agencies’ other activities, the law also applies to other data. For higher education institution libraries, the limitation to research data only applies if the library has been tasked to carry out or fund research, or to make research data directly accessible. Some public actors are exempted from the law. This applies for example to cultural institutions other than libraries, museums and archives.
Research data covered by the legislation
According to the legislation, data is defined as information in digital format, irrespective of the medium. Research data refers in the law to data that to some extent is publicly funded, collected or produced within the framework for scientific research activities, and made directly accessible for further use via a publicly accessible data platform. Research data may, for example, include statistics and results from experiments, metadata, requirement specification and other digital objects
The fact that research data is to some extent publicly funded can, for example, mean that they have been produced within the framework for a research project that is wholly or partly publicly funded. Research projects that are co-funded by a public body and a private company shall also be regarded as partly publicly funded.
Some data are excluded from the legislation. These are data that some persons, due to their identity or position, have a specific right to access according to some other law or ordinance. Nor does the law apply to data covered by sole rights, for example according to patent legislation (Patentlag (1967:837)), or that a third party has rights over according to copyright legislation (Lag (1960:729) om upphovsrätt till litterära och konstnärliga verk).
When the legislation applies
The legislation shall apply when someone who has the right to access data according to some other law or ordinance requests that data is made accessible for further use. It shall also apply when a public actor, on their own initiative, made data covered by the legislation accessible for the purpose of it being further used.
The legislation does not apply when data is transferred between national and municipal public agencies, or when data is transferred from a publicly governed body that is treated as equal to a public agency, or from a publicly owned company to a national or municipal public agency. Nor does the legislation apply when a national or municipal public agency provides data in an activity that is open to competition.
According to the legislation, research data has been made directly accessible when they are accessible to the general public for further use via a data platform. This means that the legislation does not apply for research data that is made accessible following an individual request according to national regulations governing access to data.
Preconditions for accessibility to research data
Research data shall be made accessible to the extent that requirements for data security and protection of personal data can be maintained, and on condition that they do not entail a risk to Sweden’s security. Special requirements apply in relation to format and conditions for further use.
Information about data that has been made accessible or is searchable
The Agency for Digital Government (DIGG), has been tasked to publish digitally a list of data that have been made accessible or searchable on the internet according to the legislation, and will inform about more detailed regulations for the content and design of the list, and whether the information to be provided about data that is made accessible or searchable in the list.
Actors covered by the new legislation shall provide information to DIGG about data that they, at their own initiative, have made accessible or searchable on the internet for the purpose of the data being further used.
Fees
A fee for making data accessible for further use may be charged under certain circumstances. Research data shall, however, be made accessible free of charge.
Questions about how the legislation shall be interpreted or implemented
If you are a researcher or handle research data in some other role and have a question about how the law should be interpreted or implemented in a certain case, please contact the person responsible for issues relating to accessibility of data in your organisation.
We have based the above information on the following texts:
The Open Data Act, The Authority for Digital Management website
External link.
If you are conducting research at a Swedish governmental higher education institution, your work is covered by the research principal’s* requirements for archiving and screening. It is important that you contact your research principal to learn about the guidelines and procedures for archiving and sifting that apply in your organisation. Other research principals outside the public sector may also be covered by the requirements above.
Research data shall be archived at the public agency, such as a higher education institution, where the research has been conducted. The archive personnel, data protection officer or lawyers in your organisation can provide help on issues of how to go about this. Most higher education institutions have personnel who can provide both practical help and advice.
Storage of personal data
There are rules that may require you to sift out personal data already while your research project is in progress. For example, personal data may only be processed for as long as necessary to fulfil the purpose of the processing. Personal data that is no longer needed shall therefore be removed, in a way that ensures regulations and archiving and sifting requirements are complied with and fulfilled.
What is the difference between ‘storage’ and ‘long-term preservation’?
Storage means the technical storage of data. Long-term preservation means that you also ensure that data can be found and understood over time, for example by adding relevant metadata.
*The research principal is the physical or legal entity within whose organisation the research is conducted, such as a higher education institution, a municipality, a public agency or a private company.
The teacher exclusion clause is an exclusion from the right to employees’ inventions as stated in Swedish legislation on rights to inventions at work (SFS 1949:345). Research data is normally not covered by the teacher exclusion clause. To find out what applies for specific research data, please contact your higher education institution.
No, the fact that research data are official documents does not mean the same thing as open access, that is, that research data can be published freely available on the internet.
What does ‘FAIR research data’ mean?
FAIR is an internationally recognised concept, consisting of 15 guiding principles for research data and data management. According to the FAIR principles, research data shall be
- Findable
- Accessible
- Interoperable
- Reusable
In May 2023, Sweden and the EU’s member states adopted the Council of the European Union’s conclusions that encourage member states to update their national strategies and guidelines for open access to make academic publications accessible immediately with open licences, and to apply the FAIR principles.
Criteria for FAIR research data
The Swedish Research Council has developed criteria for all 15 FAIR principles. The FAIR criteria aim to increase the understanding and knowledge of what the FAIR principles mean in practice. They can be used by researchers, research funding bodies, or for strategic decisions when implementing FAIR in the policies of organisations.
Certain aspects of FAIR are more demanding and complex to fulfil than others – and to assess the fulfilment of. Certain principles assume that other principles have been fulfilled, for example, PID (Principle 1) is a fundamental component that must be in place to enable other principles to be fulfilled. The fulfilment of certain principles entail collaboration with various support functions.
A criterion can be fulfilled in different ways, and to different degrees. Certain aspects may, in some situations, be achievable only for metadata and not for data, for example when they consist of sensitive data.
The criteria are worded in a general way, and any comparison of FAIR-ness must take into account the research field and type of data.
To support the work with FAIR data management, there is also a need for supporting infrastructure, services, and not least competence in relation to and a culture of FAIR and good data management.
Research data shall be findable
(Meta)data are assigned a globally unique and persistent identifier.
Digital data need to be given globally unique and persistent code strings that identify the object.
Persistent identifiers need to be designed in a way that enables digital objects published on the internet to be persistent and linkable with other types of digital objects.
What inputs are needed to fulfil the criterion?
Persistent identifiers shall be designed according to the applicable recommendations. In case the identifiers are only used internally to begin with, they should be designed in such a way that they can be integrated according to applicable recommendations when used online.
Data are described with rich metadata.
There should be specifically produced machine-readable metadata designed in a way that enables data to be searched.
To make research data easy to find and use for users, a brief description of these data needs to be produced, in a way that allows searches to be processed mechanically.
What inputs are needed to fulfil the criterion?
In addition to the identifier, machine-readable metadata should also include other types of descriptions, such as content, the actor publishing the digital object, the time period it was collected or created, the subject area involved, and so on.
(Meta)data are associated with detailed provenance.
The digital object’s provenance includes details such as the origin of the information, the actors who have created and/or changed the information, the activities carried out, and also in what order and when each activity was carried out.
To ensure users of the digital object can understand the purpose for which the information was produced, and by whom, the object’s provenance needs to be described.
What inputs are needed to fulfil the criterion?
The digital object’s provenance should be described in a standardised way, according to applicable recommendations.
Metadata clearly and explicitly include the identifier of the data it describes.
Metadata that describe data include a reference to a globally unique and persistent identifier for the object (the data) it describes.
Metadata should include a reference to identifiers for the object they describe. This is important to safeguard the permanence of the digital object in cases where metadata and data are made accessible separately.
What inputs are needed to fulfil the criterion?
You should strive to ensure that metadata and data can be accessed via the same persistent identifier. Machine-readable metadata should include a reference to the identifier for the object they describe.
(Meta)data are registered or indexed in a searchable resource.
Metadata that describe the digital object are searchable and simple to find on the internet. Metadata that describe a digital object should be designed to increase the visibility of the object, and to increase the potential for its use.
What inputs are needed to fulfil the criterion?
Metadata should be designed in a way that enables accessibility via one or several services, portals or repositories intended for this purpose. This can be done, for example, via machine-readable standardised metadata that permit harvesting of metadata, indexing in general search engines, etc.
Research data shall be accessible
(Meta)data are retrievable by their identifier using a standardised communications protocol.
Metadata and, if applicable, data, can be reached, read and accessed via a standardised communications protocol (such as http or ftp) through an allocated identifier. If this is not possible for some reason, for example security reasons, a description of the procedure to be used must be attached.
The technical protocol that controls the reading of digital objects on the internet should be standardised and allow the objects to be reached via their identifiers. If the objects cannot be reached for reading, or access cannot be given in an automated way, the procedure to be used must be clearly described.
What inputs are needed to fulfil the criterion?
There should be an identifier that gives access to the digital object via a standardised communications protocol. If access cannot be given via an automated protocol, the metadata should include a description of how to get access.
The protocol is open, free, and universally implementable.
The communications protocol used to reach and read the objects should be open, free and universally implementable.
The technical protocol used to read the digital objects on the internet, or to give access to the objects, should not in itself be a limiting factor for access.
What inputs are needed to fulfil the criterion?
Communication protocols used to reach the digital objects should be designed in a way that does not complicate or limit access to and reading of digital objects intended to be openly accessible.
The protocol allows for an authentication and authorisation procedure, where necessary.
The technical protocol used to read the digital objects on the internet, or to give access to the objects, should if necessary be designed in a way that allows verification that a user is who they claim to be, or authentication, and also allows limiting access to the data that the user has the right to access, or authorisation.
If there is a need for mechanisms for user roles and identification, these should be enabled through protocols that control access to the information.
What inputs are needed to fulfil the criterion?
Protocols used to reach the digital objects should be designed in a way that enables implementation of authorisation and authentication mechanisms.
Metadata are accessible, even when the data are no longer available.
Metadata that describe a digital object remain even if access to the object has ended.
Even if access to a digital object is limited or ends, access to the overarching metadata should remain in place. Some data can, for example, only be reached in a physical location, and not via the internet, but to find these data, access to metadata on the internet is still needed.
What inputs are needed to fulfil the criterion?
Metadata should be managed and made accessible over time, even when data linked to these descriptions is no longer accessible (for example via the internet).
Research data shall be interoperable
(Meta)data use a formal, accessible, shared, and broadly applicable language for knowledge representation.
To increase the potential of the use of information over time, methods for designing semantic descriptions of the information need to be standardised, documented and accessible.
To ensure users can understand how the information has been structured, they need to have access to the information, but also to the method for interpreting the information.
What inputs are needed to fulfil the criterion?
To ensure users can understand how the information has been structured, methods for designing semantic descriptions of the information need to be documented and accessible. If existing (standardised) semantic descriptions of the digital information are used, they should be described. If no standardised descriptions exist, and new descriptions are produced in conjunction with the information being created, these should be documented and made accessible together with the information.
(Meta)data use vocabularies that follow FAIR principles.
To describe the contents of digital objects, checked vocabularies, terminologies or ontologies are used that in themselves fulfil the FAIR criteria. Descriptions of these are documented and accessible.
Checked and generally accepted vocabularies, terminologies or ontologies facilitate the interpretation, linking, interoperability and reuse of data, and minimise the risk of misinterpretation of concepts and definitions.
What inputs are needed to fulfil the criterion?
Existing standardised vocabularies, terminologies or ontologies should be used. If there are none, and new ones are developed, then these should be in accordance with established standards, and should also be documented and made accessible.
(Meta)data include qualified references to other (meta)data.
Relationships between the different parts of an interrelated digital object, their meaning and their relationships with any other digital objects are described in a standardised and contextualised way.
Description of the relationships between different objects (such as whether the data in question are based on other data, or whether the data in question need to be supplemented with other information) facilitates understanding and reuse. Expressing relationships in a standardised way facilitates linking of different data and metadata.
What inputs are needed to fulfil the criterion?
Relationships between different objects should be described. To express relationships, we should strive to use generally accepted formats for this that permit linking (such as RDF, Resource Description Framework).
Research data shall be reusable
(Meta)data are richly described with a plurality of accurate and relevant attributes.
The provenance and contents of the digital objects are described using contextual metadata.
By describing the content using metadata and semantics that are as rich as possible, good preconditions are provided to ensure the contents are searchable, storable, usable and reusable, and that characteristics that affect whether data are suitable for the purpose of the reuse are discovered and can be managed as early as possible in the process.
What inputs are needed to fulfil the criterion?
The digital object’s benefit in use needs to be optimised by adding contextual metadata that describe its contents and importance, and also the context in which it was created/collected.
(Meta)data are released with clear and accessible data usage license.
The digital object shall include clear rights marking/licence. Conditions should cover both metadata and (where applicable) data.
If it is not clear in what way and on what conditions a digital object may be used, this can affect the access to and use of the information. It should therefore be clearly stated what rights markings, licences or rules that apply, for both data and metadata.
What inputs are needed to fulfil the criterion?
Information generated through scientific research needs to be made accessible in such a way that the conditions for use and reuse are clear.
(Meta)data meet domain-relevant community standards.
Standardised ways are used to describe a specific object and its various parts, which creates opportunities for understanding and interoperability.
Using standardised ways of describing domain-specific information facilitates the information being searched for, interpreted and linked.
What inputs are needed to fulfil the criterion?
The contents of digital objects should be structured and described in as standardised a way as possible, and in accordance with applicable recommendations.
Guide to implementing the criteria
To enable FAIR data management to be implemented in the research process, the Swedish Research Council has produced a more detailed description in the form of a guide to the criteria. The guide aims to provide support for practical implementation of the criteria at a more detailed level. It can be used as a joint source of information for all who support researchers in planning data management, in particular the support functions for data management at higher education institutions and infrastructures, where adaptation to FAIR data management and practical implementation of the FAIR principles are often done.
Guide to implementing the criteria for FAIR research data
Guide to implementing the criteria for FAIR research data (in Swedish)
PublISHED ON
UpDATED ON