FAIR Data by Design: A Case of the DiVA Portal

Phub Namgay

and Joshua C. Nwokeji

Department of Informatics and Media, Uppsala University, Sweden

Computer and Information Science Department, Gannon University, U.S.A.

Keywords: FAIR Data Principles, FAIR Data, FAIR Data Repositories, DiVA Portal.

Abstract: FAIR Data Principles is a guideline for making data and other digital objects findable, accessible,

interoperable, and reusable. Thus far, the traction and uptake of the principle are primarily in the domain of

bio and natural sciences. The knowledge gap is the application of the FAIR Data Principles in designing data

repositories for FAIR data in the academic data ecosystem. This paper provides a critical insight into how the

principle can be utilised as a paradigm to design data that embodies the tenets of FAIR Data Principles. We

conducted a case study of the DiVA portal, an information repository and finding tool in Sweden, to explicate

FAIR data by design. The portal scored high in a qualitative assessment against the 15 facets of FAIR Data

Principles, as illustrated by the high density of green cells in the traffic light rating matrix (see Table 1). It

indicates the robustness of data in the portal that is easy to share, find, and reuse. This study suggests

practitioners operationalise FAIR Data Principles in their data repositories by design through systems and

policies underpinned by the principle. It would enrich data governance and management for the back office

and data experiences for end users. The study also advances the knowledge base on data management through

a granular exposition of FAIR data by design.

1 INTRODUCTION

The heart of a data repository is that end users can

effortlessly share, find, and reuse data. The pervasive

datafication (Lycett, 2013) of society further

intensifies a need for robust data repositories to share

and consume data. For example, in academia,

students and professors create and share data in

information portals and data repositories for reuse by

others. In the current study, ‘data’ refers to digital

objects such as publications, theses, data sets, and

metadata. Data is also increasingly ‘born digital’ in

the academic data ecosystem. However, the lack of

coherent policies on data infrastructure to govern and

manage data impedes the processes and practices

around data. A paradigm that underpins data

repositories and data therein ‘by design’—that is, ‘by

plan’—is crucial for producing and consuming

findable and reusable data in a data ecosystem.

FAIR Data Principles is a guideline for managing

and stewarding data and other digital objects

(Wilkinson et al., 2016). It has witnessed rapid

https://orcid.org/0000-0001-6034-7274

https://orcid.org/0000-0003-4643-2418

traction in research and practice (van Reisen et al.,

2020). Nevertheless, the knowledge gap is an account

of the application of FAIR Data Principles in the

context of everyday academic data repositories to

illustrate the practicality of the principle. A study that

examines data repositories by anchoring on FAIR

Data Principles is essential, which entails scrutinising

digital objects for FAIRness (Jacobsen et al., 2020;

Wilkinson et al., 2019). DiVA portal, an information

repository and finding tool in Sweden (The DiVA

Consortium, 2022), was sampled for a case study to

pursue the research question— Why should data

repositories be FAIR by design to facilitate findable

and reusable data?

The exponential growth of data and increased

demand for data have informed us that simply sharing

data on the data repositories is insufficient today.

Reinforcing data infrastructures with data paradigms

is essential for sharing and reusing data. DiVA portal

was assessed against FAIR Data Principles

(Wilkinson et al., 2016). The portal scored high in

overall FAIRness, as substantiated by many green

cells in a colour-coded matrix using a traffic light

160

Namgay, P. and Nwokeji, J.

FAIR Data by Design: A Case of the DiVA Portal.

DOI: 10.5220/0011964300003467

In Proceedings of the 25th International Conference on Enterprise Information Systems (ICEIS 2023) - Volume 2, pages 160-167

ISBN: 978-989-758-648-4; ISSN: 2184-4992

 2023 by SCITEPRESS – Science and Technology Publications, Lda. Under CC license (CC BY-NC-ND 4.0)

rating (Dunning et al., 2017). This study posits that

data repositories, such as those used daily in the

academic data ecosystem, must consider and

operationalise FAIR Data by design to facilitate data

governance and management for the back office and

a seamless data experience for end users.

In the following sections, first, we provide an

overview of FAIR Data Principles. Then, an account

of the study design and method is presented to set the

tone, followed by the analysis and findings of the

study. Next, we discuss the interpretation of the

findings, contribution to knowledge and practice, and

study limitations. The paper finally ends with a

conclusion of the study.

2 OVERVIEW OF FAIR DATA

PRINCIPLES

FAIR Data Principles is a guideline for making data

and other digital objects findable, accessible,

interoperable, and reusable (Wilkinson et al., 2016),

as in Figure 1. The principle emphasises data reuse

(David et al., 2020) and offers a middle ground for

sharing and reusing data on the spectrum of ‘close

data’ and ‘open data’ (Boeckhout et al., 2018).

FAIR Data Principles have gained traction and

uptake in the research, government, and practice

(Jacobsen et al., 2020; van Reisen et al., 2020;

Wilkinson et al., 2019). The H2020 Program

Guideline underlines FAIR Data Principles by

highlighting that data should be “as open as possible,

as closed as necessary” (Celina, 2017; Landi et al.,

2020). The inequality in implementing FAIR Data

Principles across different world geographies is a

concern as high uptake is reported in the Western and

Northern hemispheres (van Reisen et al., 2020),

predominantly in the domain of bio and natural

sciences (Dunning et al., 2017; van Reisen et al.,

2020).

The facets of FAIR Data Principles are concise

and thus susceptible to diverse interpretations (David

et al., 2020; Jacobsen et al., 2020). However, the

interpretations should still be within the spirit of the

original facets or established behavioural guidelines

of the principle (Mons et al., 2017), even if the

requirements for FAIR data vary across disciplines

(Dunning et al., 2017). The unqualified application of

FAIR Data Principles also risks creating additional

issues for data stewardship (Boeckhout et al., 2018).

FAIR Data Principles has not been seriously used

in everyday data repositories or digital artefacts, such

as those used daily in academia. The discourse and

literature in this area are scant. Thus, it is in order to

explore how data repositories in the academic data

ecosystem operationalise FAIR Data Principles and

explicate FAIR data by design to ease the production

and consumption of data.

Figure 1: FAIR Data Principles and its facets (Wilkinson et

al., 2016).

3 STUDY DESIGN AND METHOD

3.1 Study Context

DiVA portal is an institutional information repository

or publishing database established for research,

digital materials, and thesis publications by the 50

member universities, colleges, institutes, and

museums of the DiVA consortium in Sweden (The

DiVA Consortium, 2022). It is a common tool for

finding and reusing academic resources, non-

academic materials, and data sets. DiVA portal

currently stores about 20 different types of

publications, including data sets. It also provides a

service for the long-term archiving of scholarly

output. The portal promotes Open Science, and digital

objects therein are globally accessible.

3.2 Study Method

A case study was conducted to examine a data

repository against the four foundational principles

and 15 facets of FAIR Data Principles, as in Figure 1.

DiVA portal was sampled as a case to answer the

FAIR Data by Design: A Case of the DiVA Portal

161

research question. The portal is a widely used

information system in Sweden. Furthermore, with the

hope of disseminating research, findings, and

knowledge, the portal is a dynamic system with

continual new data every day. It is also apparent from

using the portal that it would be compelling to explore

how the portal embodies the facets of FAIR Data

Principles. We also gathered data about the portal,

through email correspondence, from a university

library and the data office of one of the universities in

the DiVA consortium. Their experiential insights and

perspectives on FAIR data were crucial for the in-

depth analysis of the portal.

4 ANALYSIS AND FINDINGS

The following subsections illustrate the manual

assessment of the DiVA portal against the 15 facets

of FAIR Data Principles (see Dunning et al., 2017).

In the interest of space, we layered the figures as the

text behind the inset is not central in the current study.

4.1 Findable Data

The first task in locating data of interest is to find it in

a data repository. Metadata, defined as ‘data about

data’, should facilitate finding data, and machine-

actionable metadata is desired for the automatic data

search.

F1. (Meta)data are assigned a globally unique

and persistent identifier—DiVA portal uses unique

persistent identifier (PID), such as Uniform Resource

Name (URN), Open Archive Initiative OAI, DiVA id,

and Digital Object Identifier (DOI) to identify data on

the portal uniquely. The portal automatically mints a

URN: NBN (Uniform Resource Name for National

Bibliography Number) while registering any digital

objects on the portal, as shown in Figure 2.

F2. Data are described with rich metadata— All

the digital objects on the DiVA portal are described

with rich metadata for meaningful reuse, as shown in

Figure 2. Most of the information in metadata is

captured while registering data on the portal.

F3. Metadata clearly and explicitly include the

identifier of the data it describe—The PID of data,

that is, URN, is captured in the metadata, as shown in

Figure 3. Metadata is viewable through the portal’s

‘Export’ feature, which exports metadata in 13

different formats. For example, the Electronic Theses

and Dissertations-Metadata Standard (ETD-MS) is a

metadata standard for electronic theses and

dissertations based on Dublin Core.

F4. (Meta)data are registered or indexed in a

searchable resource—Data and its associated

metadata are registered in a searchable Web-based

resource, as shown in Figure 3 (see Figure 2 for PID).

It is also consistent with linking data and metadata

(Jacobsen et al., 2020).

Figure 2: Facet F1 and F2 of FAIR Data Principles.

ICEIS 2023 - 25th International Conference on Enterprise Information Systems

162

Figure 3: Facet F3 and F4 of FAIR Data Principles.

Figure 4: Facet A1, A1.1, A 1.2 (inset), and A2 of FAIR Data Principles.

4.2 Accessible Data

Once the data is found, a user needs to know how it

can be accessed using authentication and

authorisation mechanisms (Landi et al., 2020), such

as gatekeeper, open-access, or a mix of both. This

information is necessary to ensure data can be located

and reused, not simply found and viewed on a data

repository, through well-defined data ‘access’

control.

FAIR Data by Design: A Case of the DiVA Portal

163

A1. (Meta)data are retrievable by their identifier

using a standardised communications protocol—

DiVA portal uses HTTPS communication protocol to

deliver data on the Web, as in Figure 4. The

associated metadata is also retrievable from the Web.

It is worth noting that data can be found yet not

accessible.

A1.1 The protocol is open, free, and universally

implementable—DiVA portal is open and accessible

through an HTTPS connection, as in Figure 4.

A1.2 The protocol allows for an authentication

and authorisation procedure, where necessary—A

user must log in using the university or institute login

credential or Single sign-on to share any data on the

DiVA portal, as in Figure 4. It resonates with the

controlled access model (Landi et al., 2020) for a

regulatory and ethical check. Any data to be shared

on the portal must be bibliographically approved.

A2. Metadata are accessible, even when the data

are no longer available— Metadata of data on the

DiVA portal is accessible even if data ceases to exist.

Personnel from the Scholarly Communication

Division of one of the universities in the DiVA

consortium noted: “Yes, metadata, i.e. the description

of the data, will still be available in DiVA even if the

data no longer exists. Once registered in DiVA, the

record with the metadata will receive a permanent

identifier and link. The only way the metadata can

disappear from DiVA is if someone hides or deletes

the record itself.”

4.3 Interoperable Data

Data is often interoperated or integrated with other

data, workflows, and applications. It can also be an

input to workflow systems.

I1. (Meta)data use a formal, accessible, shared,

and broadly applicable language for knowledge

representation— Different metadata schemas are

used for organising knowledge. DiVA portal uses the

Metadata Object Description Schema (MODS), a

bibliographic element set schema, as shown in Figure

5. The information captured in the portal for

registering data is based on the DataCite metadata

schema. The portal also provides the feature to

retrieve digital e-archive files in Metadata Encoding

and Transmission Standard (METS) format.

I2. (Meta)data use vocabularies that follow FAIR

principles— This facet was challenging to assess.

Dunning et al. (2017) reported a similar experience.

However, personnel from the data office of one of the

universities in the DiVA consortium shared: “DiVA

developers are doing some other FAIR-related work

in DiVA, like making it possible to register licences in

the form.”

I3. (Meta)data include qualified references to

other (meta)data—‘Qualified references’ refers to

informational links that would help disambiguate the

terms used (Dunning et al., 2017). Data shared on the

DiVA portal has a qualified reference to the

publication, as in Figure 5. The PID of the research

publication is linked to its data set through ‘The data

set is referenced by’ and ‘The publication has

references to’ features in the DiVA Portal.

4.4 Reusable Data

The goal of sharing data on various data repositories

is to reuse it in order to create an innovative solution

or solve problems.

R1. Meta(data) are richly described with a

plurality of accurate and relevant attributes—Data

on the DiVA portal is described with rich metadata.

For example, in the context of a data set, the ETD-MS

includes essentials such as PID, authors, and

provenance, as shown in Figure 6. The metadata for

publications includes richer information on the

business, technical, and operational aspects, such as

identifiers, conferences, journals, and publishers, as

in Figure 5 (see inset).

R1.1. (Meta)data are released with a clear and

accessible data usage license—DiVA portal

mentions the publishing conditions. The portal clearly

states that the copyright of any data belongs to the

author, as shown in Figure 6. Data on the portal is

open access. Likewise, the portal recommends

authors share their work with a licence such as

Creative Commons.

R1.2. (Meta)data are associated with detailed

provenance—Metadata captures information on the

provenance of the data. For example, information

such as author, publisher, and identifier is captured

across different metadata schemas, as shown in

Figure 6.

R1.3. (Meta)data meet domain-relevant

community standards—Unlike other facets, which

are observable by brief investigations such as facet

‘F1’, some facets require detailed subject knowledge,

such as assessing whether data and its associated

metadata meet community or discipline-specific

standards. Examining whether the data shared on the

DiVA portal explicitly meets domain-specific

metadata and data standards is difficult. Dunning et

al. (2017) found some facets challenging to assess and

gave the lowest rating—‘unclear’ in traffic light

rating.

ICEIS 2023 - 25th International Conference on Enterprise Information Systems

164

Figure 5: Facet I1, I2, and I3 of FAIR Data Principles.

Figure 6: Facet R1, R1.1, R1.2, and R1.3 of FAIR Data Principles.

FAIR Data by Design: A Case of the DiVA Portal

165

5 DISCUSSION

The following subsections discuss the interpretation

of the findings, contributions to knowledge, and

limitations of the study.

5.1 FAIR Data by Design

A fundamental quality of data repositories is the ease

of searching, finding, and reusing data. It is evident

from the findings that data repositories underpinned

by FAIR Data Principles would enrich the processes

and practices of sharing and consuming data via

coherent PIDs, standards, metadata, and provenance.

Thus far, the principle has not been explicitly applied

to study data repositories used daily in academia (see

van Reisen et al., 2020). We posit that everyday data,

such as data in data repositories or information portals

of the academic data ecosystem, should be FAIR by

design to ease finding and reusing data for end users.

The operationalisation of FAIR Data Principles

would enrich the governance and management of

data. Data repositories could mandate data to be FAIR

explicitly through policies and frameworks or

implicitly by implementing facets of FAIR Data

Principles in information systems by design—that is,

data repositories implement FAIRification features

(Jacobsen et al., 2020). It will encourage those who

want to share and reuse data to consider FAIR Data

Principles, and eventually, the principle will become

second nature in routine data practices.

FAIR Data by design necessitates addressing

some challenges. Some vague facets of FAIR Data

Principles are difficult to assess in practice, even with

adequate knowledge (see Dunning et al. 2017). How

does one define ‘rich’ metadata for digital objects? A

situation might demand specificity and granularity in

the metadata of data. In contrast, it is not an issue with

facets that only require a brief examination of visible

attributes in data, such as PIDs or indexed in a

searchable resource. In sum, data practitioners should

operationalise FAIR Data Principles according to

one’s requirements in data.

5.2 FAIR Data Repositories

A data repository that simply gathers, integrates, and

stores data is insufficient in today’s complex

sociotechnical systems. It is essential to have data

repositories that bear the hallmark of FAIR Data

Principles for dealing with a wide assortment of data.

Furthermore, as FAIR Data Principles gains traction

and uptake in various areas (van Reisen et al., 2020),

it will likely inform the paradigm to share and reuse

data. Hence, know-how and skills to operationalise

FAIR Data Principles, especially by manual means,

are crucial for all stakeholders (David et al., 2020).

Moreover, a proactive approach to FAIRify data

repositories will pay dividends through seamless data

governance and management in a data ecosystem.

Table 1 illustrates the traffic-light rating matrix

(Dunning et al., 2017) for a quick summary of the

FAIRness score of the DiVA portal. One could strive

to maximise the score; nonetheless, it should not be a

goal in and of itself. The score does not have an

intrinsic value. Rather, the score should be used

meaningfully to gain insights and knowledge about a

data repository to facilitate a smooth data experience

for the back office and end users. The assessment of

a data repository for FAIRness is not a binary test of

‘Pass/FAIR’ or ‘FAIL/Not FAIR’. If a data repository

scores low in FAIRness, it provides an avenue to

improve a data repository. The manifestation of the

characteristics of FAIR in a particular context is open

to subjective interpretation (Mons et al., 2017).

Sometimes a repository may perform poorly in the

FAIRness due to various factors, such as data being

accessible on the Web or other mediums, but

insufficiently described (van Reisen et al., 2020) or

only using basic README files. It is imperative to

consider the context of what needs to be a FAIR data

repository.

Table 1: Traffic Light Rating Matrix of the FAIRness score

of the DiVA portal.

FAIR Data

Principles

15 Facets of FAIR Data Principles

Green [G]: compliant, orange [O]: just

about/maybe not, red [R]: not

compliant, and blue [ B]: unclear

Findable F1

[G]

Accessible A1

[G]

A1.1

[G]

A1.2

[G]

Interoperable I1

[G]

[O]

[G]

Reusable R1

[G]

R1.1

[G]

R1.2

[G]

R1.3

[O]

5.3 Contribution to Knowledge and

Limitations of the Study

This study advances the knowledge base of enterprise

data management from a fresh perspective, namely

the granular exposition of FAIR data by design

through a case study. In addition, the study

contributes to the literature on FAIR Data Principles

by expanding its application in other fields (van

Reisen et al., 2020). This study suggests practitioners

ICEIS 2023 - 25th International Conference on Enterprise Information Systems

166

implement FAIR Data Principles in their data

repositories through data policies underpinned by the

principle and embed facets of the principle in the

system by design. It would enrich data governance

and management for the back office and create

seamless data experiences for end users.

Dunning et al. (2017) highlighted the concise

nature of the facets of FAIR Data Principles. The

possibility of various interpretations is inevitable.

Hence, the reproducibility of this study should yield

results within an acceptable level of variation of the

frame of the principle. This study is qualitative, and

we suggest quantitative or mixed methods studies to

further develop the concept of FAIR data by design.

6 CONCLUSION

A case study of the DiVA portal in Sweden was

conducted to explicate FAIR data by design. This

study advances the knowledge base of data

management through an in-depth exposition of

operationalising FAIR Data Principles in designing

information repositories for FAIR data. This study

suggests practitioners consider implementing FAIR

Data Principles as a data paradigm that underpins

their data repositories through policies that

underscore the principle and implement facets of the

principle in the system by design. Data sources that

implement the principle facilitate the production and

consumption of findable and reusable data for the end

users. In addition, this study acquaints practitioners

with a manual assessment of data repositories of

interest against FAIR Data Principles and the

challenges of realising the principle in practice. In

sum, FAIR data by design is a way forward to govern

and manage data in a dynamic data ecosystem.

REFERENCES

Boeckhout, M., Zielhuis, G. A., & Bredenoord, A. L.

(2018). The FAIR guiding principles for data

stewardship: fair enough? European Journal of Human

Genetics, 26(7), 931-936. Retrieved from

https://doi.org/10.1038/s41431-018-0160-0

Celina, R. (2017). H2020 Programme-Guidelines on FAIR

Data Management in Horizon 2020. Retrieved from

https://policycommons.net/artifacts/1940350/h2020-

programme/2692119/

David, R., Mabile, L., Specht, A., Stryeck, S., Thomsen,

M., Yahia, M., & Bailo, D. (2020). FAIRness Literacy:

The Achilles’ heel of applying FAIR principles.

CODATA Data Science Journal, 19(32), 1-11.

Retrieved from http://doi.org/10.5334/dsj-2020-032

Dunning, A., De Smaele, M., & Böhmer, J. (2017). Are the

FAIR Data Principles fair? International Journal of

Digital Curation, 12(2), 177-195. Retrieved from

https://doi.org/10.2218/ijdc.v12i2.567

Jacobsen, A., de Miranda Azevedo, R., Juty, N., Batista, D.,

Coles, S., Cornet, R., & Evelo, C. T. (2020). FAIR

principles: interpretations and implementation

considerations. Data Intelligence, 2(1-2), 10-29.

Retrieved from https://doi.org/10.1162/dint_r_00024

Jacobsen, A., Kaliyaperumal, R., da Silva Santos, L. O. B.,

Mons, B., Schultes, E., Roos, M., & Thompson, M.

(2020). A generic workflow for the data FAIRification

process. Data Intelligence, 2(1-2), 56-65. Retrieved

from https://doi.org/10.1162/dint_a_00028

Landi, A., Thompson, M., Giannuzzi, V., Bonifazi, F.,

Labastida, I., da Silva Santos, L. O. B., & Roos, M.

(2020). The “A” of FAIR–as open as possible, as closed

as necessary. Data Intelligence, 2(1-2), 47-55.

Retrieved from https://doi.org/10.1162/dint_a_00027

Lycett, M. (2013). ‘Datafication’: Making sense of (big)

data in a complex world. European Journal of

Information Systems, 22(4). Retrieved from

https://doi.org/10.1057/ejis.2013.10

Mons, B., Neylon, C., Velterop, J., Dumontier, M., da Silva

Santos, L. O. B., & Wilkinson, M. D. (2017). Cloudy,

increasingly FAIR; revisiting the FAIR Data guiding

principles for the European Open Science Cloud.

Information Services & Use, 37(1), 49-56.

doi:10.3233/ISU-170824

The DiVA Consortium. (2022). About DiVA. Retrieved

from https://www.info.diva-portal.org/about-diva/

van Reisen, M., Stokmans, M., Basajja, M., Ong'ayo, A. O.,

Kirkpatrick, C., & Mons, B. (2020). Towards the

tipping point for FAIR implementation. Data

Intelligence, 2(1-2), 264-275. Retrieved from

https://doi.org/10.1162/dint_a_00049

Wilkinson, M. D., Dumontier, M., Aalbersberg, I. J.,

Appleton, G., Axton, M., Baak, A., & Bourne, P. E.

(2016). The FAIR Guiding Principles for scientific data

management and stewardship. Scientific Data, 3(1), 1-

9. doi:10.1038/sdata.2016.18

Wilkinson, M. D., Dumontier, M., Sansone, S.-A., Bonino

da Silva Santos, L. O., Prieto, M., Batista, D., & Crosas,

M. (2019). Evaluating FAIR maturity through a

scalable, automated, community-governed framework.

Scientific Data, 6(1), 1-12. Retrieved from

https://doi.org/10.1038/s41597-019-0184-5

FAIR Data by Design: A Case of the DiVA Portal

167