Specification of Personal Data Protection Requirements
Analysis of Legal Requirements from the GDPR Regulation
Mário Fernandes
1
, Alberto Rodrigues
da Silva
1
and António Gonçalves
1,2
1
INESC-ID, Instituto Superior Técnico, Universidade de Lisboa, Portugal
2
Escola Superior de Tecnologia de Setúbal, Instituto Politécnico de Setúbal, Portugal
Keywords: Personal Data Protection, Requirements Specification, Rslingo, Regulation (EU) 2016/679, GDPR.
Abstract: The European Union establishes in the Regulation 2016/679, or GDPR (General Data Protection
Regulation), a set of legal dispositions to achieve the protection of natural persons in what personal data
processing and the free movement of such data is concerned. When those dispositions are considered in the
development of information systems, the later become attainable for legal approval within that scope. This
paper presents the methodology we are following to elaborate a reusable catalogue of personal data
protection requirements aligned with the GDPR. Following a separation-of-concerns approach, the
catalogue shall serve the purpose of constructing information systems able to communicate with those that
process individuals’ personal data, to materialize the regulatory data protection capabilities disposed in the
GDPR. In that context, the elicitation of system requirements demands for the interpretation of a legal
document by business analysts, which consists of a scientifically relevant challenge. This research is
contextualized by the RSLingo initiative, a model-driven requirements engineering approach for the
rigorous specification of system requirements. In particular this paper discusses the GDPR’s requirements
defined as a catalogue of both business goals and system goals.
1 INTRODUCTION
The need to protect valuable information,
particularly of individuals, is a societal challenge
imposed by many governments to organizations.
Within the European scope, the General Data
Protection Regulation (GDPR or Regulation
2016/679) was defined by the European Union,
consisting of a set of legal dispositions to achieve
the protection of natural persons in what personal
data processing and the free movement of such data
is concerned (EU, 2016). The guarantee of
compliance with the GDPR, at the level of
information systems, demands for an effort from
business analysts when interpreting such legal
document.
The goal of this paper is to propose a rigorous
specification of requirements included in the GDPR,
and consequently to promote a better and systematic
interpretation of those legal requirements.
From the analysis of the GDPR document, the
aim is to elaborate a reusable catalogue of personal
data protection requirements to be used in the
design and implementation of information systems
able to communicate with those that process
individuals’ personal data, to turn them attainable
for legal approval within the scope of personal data
protection, specifically within the provisions of the
GDPR.
This research is contextualized within the
RSLingo initiative, a model-driven requirements
engineering approach for the rigorous specification
of system requirements (Ferreira & Silva, 2012;
Silva, 2015). The analysis of the GDPR involved
several tasks, namely reading, manual knowledge
extraction, and caracterization of many concepts and
sentences expressed in that legal document. The
analysis was supported by the RSLingo RSL
language (Silva, 2017, 2018), in which requirements
are defined at different abstraction levels, yet they
always represent an expression of stakeholders’
needs. Stakeholder, glossary of terms, business
process and business goal constructs allow to
express a high-abstraction-level overview of
stakeholders’ needs and concerns. From those,
system goal, functional requirement, quality
requirement, constraint, use case or user story
constructs can be used to specify their concerns at
system level. From the later, test cases can be
derived and defined in order to drive a verification
398
Fernandes, M., Rodrigues Silva, A. and Gonçalves, A.
Specification of Personal Data Protection Requirements.
DOI: 10.5220/0006810603980405
In Proceedings of the 20th International Conference on Enterprise Information Systems (ICEIS 2018), pages 398-405
ISBN: 978-989-758-298-1
Copyright
c
2019 by SCITEPRESS – Science and Technology Publications, Lda. All rights reserved
process (Silva, Paiva, & Silva, 2018). At a particular
point in time, especially if the software development
project is in its earlier stages, it may be just enough
to specify these requirements using only business
and/or system goal constructs.
The paper is structured as follows. Section 2
introduces the RSLingo RSL language, which
supports this research. Section 3 explains our
methodology of systematically analysing the GDPR
based on the background research. Section 4
presents some results obtained from applying this
methodology, namely through the presentation of
some produced work products. Section 5 discusses
some challenges addressed by our research, together
with its main contribution, and critically elaborates
on related work. Finally, Section 6 provides for
some concluding remarks and future direction for
our research.
2 BACKGROUND
RSLingo is a long-term research initiative in the
Requirements Engineering area (Ferreira & Silva,
2012). It is a linguistic approach to improve the
quality of requirements specifications. Although
being the most common and preferred form of
representing requirements, natural language is prone
to producing ambiguous and inconsistent
documents, hard to automatically validate or
transform into other kinds of artefacts (Silva, 2017).
Figure 1: RSL metamodel (simplified). The constructs in dark grey background are applied in the scope of this paper.
Businesslevel
Systemrelationsview
Systemlevel
Systemrequireme nt sview
Sta tema ch i nesview
Systemsvie w
Entities vie w
Actors view
Businessgoals viewGlossaryview
Businessprocessesview Stakeholdersvie w
BusinessProcess
GlossaryTerm BusinessGoal
Actor
UseCase
DataEntity
Stakeholder
FunctionalRequirement
DataEntityVie w
QualityRequire m ent
Constraint
StateMac hine
+
+Action
+State
SystemGoal
SystemRelation
System
UserStory
Glossary
BusinessEvent
BusinessFlow
Entity
Requirement
refere nc e
maste r
source
deta i l
initiates
pa rt ici pates
participant
externalparticipant
ta rget
Specification of Personal Data Protection Requirements
399
RSLingo RSL. (Requirements Specification
Language) (Silva, 2017, 2018) is a comprehensive
domain-specific language designed to address
general-purpose RE activities such as the rigorous
specification, automatic validation, persistence and
management of system requirements. RSL is based
on other languages such RSL-IL (Ferreira & Silva,
2013), RSL-IL4Privacy (Caramujo & Silva, 2015;
Silva, Caramujo, Monfared, Calado, & Breaux,
2016), Pohl’s (Pohl, 2010), XIS* (Ribeiro & Silva,
2014a, 2014b; Silva, Saraiva, Silva, & Martins,
2007) and SilabREQ (Savić et al., 2015). In a more
detailed way, RSL is a controlled natural language to
support the elaboration of SRSs (Systems
Requirements Specifications) in a systematic,
rigorous, consistent and less ambiguous way,
namely by using the RSL Excel Template, which is
based in RSL itself. Representing domain
knowledge with such SRSs has in itself a way of
providing business stakeholders with a better
understanding of natural language statements that
represent requirements. RSL is a language that
includes a rich set of fundamental RE-specific
constructs logically arranged into views, according
to two dimensions: abstraction levels and RE
concerns, which means the architecture of RSL is
bidimensional and multiview (Silva, 2017).
Figure 1 depicts a simplified version of the RSL
metamodel. The two abstraction levels are illustrated
in the diagram: the business level and the system
level. The business level groups the views closer to
the business perspective, whereas the system level
groups the views closer to the system perspective.
Each view gathers one or more constructs. System
constructs may depend on some business constructs
(e.g. Entities may depend on Glossary terms and
Actors may depend on Stakeholders). These
constructs are defined as linguistic patterns and
represented textually by mandatory or optional
fragments (text snippets) (Silva, 2017). The RSL is
structured according a large set of constructs, such
as: glossary term, stakeholder, business goal,
business process, business event, business flow,
system, actor, data entity and data entity view,
system goal, functional requirement, quality
requirement, constraint, use case, user story, state
machine (Silva, 2017).
The personal data protection requirements
(extracted from the GDPR) are under specification
with a set of RSL constructs selected according to
the chosen linguistic style, appropriate for the early
stages of the requirements specification. That set is
composed of some business level views together
with the system goals view. (However, we could
have specified these requirements at the system level
following other styles by using constructs such as
functional requirements, quality requirements and
constraints; use cases or user stories).
3 GDPR ANALYSIS
The GDPR is targeted at the protection of natural
persons regarding the processing of their personal
data by automated or manual means, only in the case
that those data are contained or intended to be
contained in filing systems. According to the GDPR,
a filing system is set of structured personal data
accessible via specific criteria, whether centralized
or not. However, the GDPR mentions not only filing
systems, but also (secure) systems to which
controllers may grant remote and direct access of
data subjects to their own personal data.
Figure 2: Systems and stakeholders defined in the scope of GDPR.
«System»
Processing System
(A)
«System»
Regulatory System
data subject (A)
controller
data protection
officer
processor
superv isory authority
«System»
Processing
System (B)
data subject (B)
ICEIS 2018 - 20th International Conference on Enterprise Information Systems
400
Nevertheless and based on an interpretation of
the GDPR, other systems may also be considered,
namely systems where controllers or processors
could register violations of data, where data subjects
could request for the rectification of data, where
controllers or processors could record processing
activities and where data protection officers could
monitor that registration.
For the purpose of this paper, the GDPR is an
elicitation source of not only business but also
system requirements to consider when developing
information systems able to communicate with those
that process individuals’ personal data, in order to
materialize the regulatory data protection
capabilities disposed in the GDPR. This may imply
the existence of two types of systems: regulatory
systems and processing systems, whose operation
includes, yet it is not restricted to processing
individuals’ personal data. Figure 2 shows the two
types of systems mentioned based on the GDPR‘s
Introductory items 15 and 16, and Article 2(1). This
paper presents not only business and system
requirements of regulatory systems, but also a RSL-
based methodology to extract these requirements
from legal documents. In the context of personal
data protection, system requirements state the way in
which the business requirements shall be met in the
application domain i.e. how the data shall be
protected regardless of the technology used by those
systems to implement that protection.
Figure 3: Methodology for analysing the GDPR.
1.
contextualize
2. define
stakeholders
4.1. define
systems
4.2. define
system
relations
5.1. define
business
processes
5.2. define
business
goals
6. define
system goals
Glossary
Stakeholders
Domain
model w ith
domain terms
A
System
Relations
SystemGoals
BusinessGoals
BusinessProcessesSystems
3. represent
domain terms
(and
stakeholders)
Specification of Personal Data Protection Requirements
401
The analysis of the GDPR involved a set of tasks
as depicted in Figure 3. The objects shown in this
figure represent work products as partially include in
a RSL specification.
Task 1: Contextualization. This task consisted of
the familiarization of the business analyst with the
domain of personal data protection. The work
product of this task is a list containing the definition
of the business domain terms to better support the
understanding of regulatory systems, from a
business perspective.
Task 2: Definition of Stakeholders. This task
consisted of identifying and characterizing the
parties interested in the development and operation
of regulatory systems.
Task 3: Representation of Structure Composed of
Domain Terms (and stakeholders). This task
consisted of modelling the relation between the main
domain terms identified in the GDPR, along with
their attributes. A UML class diagram can be
elaborated to complement that understanding.
Task 4: Definition of Systems and System
Relations. This task consisted of (subtask 4.1)
identifying each system involved in the business
domain of personal data protection, regardless of
being an in-scope or an out-of-scope system, and
characterizing its type e.g. whether system or
subsystem; and (subtask 4.2) characterizing the
interactions between source and target systems
involved in the business domain of personal data
protection, whether those systems being internal
systems (in-scope) or external systems (out-of-
scope).
Task 5: Definition of Business Processes and
Business Goals. This task consisted of (subtask 5.1)
identifying and characterizing the business
processes, based on an ordering of tasks concerned
with regulating the processing of personal data; and
(subtask 5.2) identifying and characterizing the
businessgoals to be achieved with the application of
the provisions in the GDPR to regulatory systems.
Although not represented in the diagram, this task
also included the identification and characterization
of business flows to express control flows between
business processes or between business processes
and business events – business flows can be of type
sequential, conditional (equivalent to the BPMN
exclusive gateway) or parallel (equivalent to the
BPMN parallel gateway).
Task 6: Definition of System Goals. This task
consisted of identifying and characterizing the
systemgoals to be achieved with the application of
the provisions in the GDPR to regulatory and
processing in-scope systems.
The execution of these tasks is guided by a
control flow that suggests the adopted GDPR
analysis methodology. Figure 3 is complemented by
the illustration of dependencies between the tasks
and the RSL constructs that are progressively
specified throughout that flow.
4 SPECIFICATION OF DATA
PROTECTION
REQUIREMENTS WITH THE
RSL
This section demonstrates part of the work products
originated from each task of the GDPR analysis
methodology described in Section 3. Some columns
from the tables in the RSL Excel Template were
omitted due to space restrictions (e.g. the
Description or Priority BusinessGoals’ attributes
were omitted in the corresponding table).
4.1 Stakeholders View
Stakeholders are people and organizations that
influence the development of an information system
or will be affected by its operation, in this case, the
operation of regulatory systems.
Table 1: Stakeholders identified by GDPR.
Name Description Type
Data
Subject
An identifiable natural person who can
be identified, directly or indirectly, in
particular by reference to an identifier
such as a name, an identification
number, location data, an online
identifier or to one or more factors
specific to the physical, physiological,
genetic, mental, economic, cultural or
social identity.
Person
Controller
A natural or legal person, public
authority, agency or other body which,
alone or jointly with others, determines
the purposes and means of the
processing of personal data
Team
Processor
A natural or legal person, public
authority, agency or other body which
processes personal data on behalf of the
controlle
r
Team
Supervisory
Authority
An independent public authority which
is established by a EU Member State.
Organi-
zation
Data
Protection
Officer
(DPO)
An enterprise security leadership role
responsible for overseeing data
protection strategy and implementation
to ensure compliance with GDPR
requirements.
Person
ICEIS 2018 - 20th International Conference on Enterprise Information Systems
402
Stakeholders are original sources of requirements,
despite that, in the case of this paper, requirements
were elicited from the GDPR itself. Table 1 contains
some stakeholders interested in the development of
regulatory systems. Some stakeholders will be users
of those systems, hence will be constituted Actors in
the Actors view. A Stakeholder is characterized, at
least, by its name and type, as stated in the table.
4.2 Business Goals View
Business goals can be faced as the value that the
system represents for the business, in this case that
regulatory system represent to the protection of
personal data and the free movement of such data.
Table 2 lists the definition of some business goals
that, once achieved with the development and
operation of regulatory systems, allow warranting
the regulatory data protection capabilities
provisioned by the GDPR. The identification and
characterization of business goals constitutes a
starting point for the identification and
characterization of system goals. The values in the
column Part of indicate the aggregation relations
that may exist between BusinessGoals, which means
that they can be decomposed into sub-
BusinessGoals. (The tags preceding the
BusinessGoals are their id(entifiers) and express no
particular order in what the items or articles of the
GDPR are concerned.)
Table 2: Some business goals extracted from the GDPR.
Name Part of
bg_1: Facilitate the exercise of the data
subject’s rights
bg_1.1: Right to obtain from the controller
confirmation as to whether or not personal
data concerning him/her are being
p
rocesse
d
bg_1
bg_1.2: Right to transmit personal data to
another controller without hindrance from
the controller to whom they have been
p
rovide
d
bg_1
bg_6: Lawfulness, fairness and
transparency of personal data processing
(
art. 6
)
bg_7: Conditions for consent (art. 7)
bg_7.1: Where processing is based on
consent, the controller shall be able to
demonstrate that the data subject has
consented to processing of his or her
p
ersonal data.
bg_7
bg_7.2: The data subject shall have the
right to withdraw his or her consent at any
time.
bg_7
4.3 System Goals View
System goals can be derived from business goals,
since the first operationalize the later in a particular
system context (in the case of this paper, the context
of the regulatory system). This means that if system
goals are met, business goals are equally met, hence
the provisions in the GDPR are respected by the
regulatory system. Table 3 depicts the definition of
some system goals extracted from the GDPR for
regulatory systems. Additionally, to allow a
comparative insight over both types of system, Table
4 presents some system goals also extracted from the
GDPR for processing systems. The semantics
associated with each column in both tables is the
same as the one associated with the column of same
name in Table 2.
Table 3: System goals extracted from the GDPR for
regulatory systems.
Name Part of
sg_1: Record personal data processing
activities
sg_1.1: Maintain a record of the personal
data processing activities under the
responsibility of each controlle
r
sg_1
sg_1.2: Maintain a record of all categories
of personal data processing activities
carried out by each processor on behalf of
a controlle
r
sg_1
sg_2: Personal data breaches shall be
notified to the competent supervisory
authority and, eventually, the data subject
Table 4: System goals extracted from the GDPR for
processing systems.
Name Part of
sg_1: Ensure a level of security
appropriate for the data subject’s rights
and freedoms
sg_1.1: Pseudonymize and encrypt
ersonal data
sg_1
sg_1.2: Keep the confidentiality, integrity,
availability and resilience of processing
systems and services
sg_1
sg_1.3: Restore the availability and access
to personal data, in a timely manner, in the
event of a physical or technical incident
sg_1
sg_1.4: Execute a process to regularly test,
assess and evaluate the effectiveness of
technical measures concerned with the
security of personal data processing
sg_1
Specification of Personal Data Protection Requirements
403
5 DISCUSSION
Analysing a legal document like the GDPR
encompasses a challenge for business analysts
working in the development of information systems
capable of processing personal data according to the
legal dispositions in that document, as well as for
business analysts working in the development of
information systems responsible for providing the
electronic means to regulate the processing of
personal data.
The analysis effort required by the parties that
must comply with the dispositions in the GDPR is
reduced with the reuse of the work products from the
analysis methodology reported in this paper. On one
hand, reporting to the business level work products
(e.g. Glossary, Stakeholders and BusinessGoals),
they support a better understanding of the GDPR,
from which regulatory systems can be further
designed and implemented. They also augment the
comprehensibility of the personal data protection
business domain. On the other hand, reporting to the
system level work products (e.g. Systems and
SystemGoals), they provide for a rigorous
interpretation of the GDPR, from which, specifically
in the scope of our research, regulatory systems can
be further designed and implemented.
Prior efforts of other authors have been
undertaken to systematize past research concerned
with the handling of legal texts for software systems
development (Otto & Antón, 2007). The same
authors who surveyed those efforts, together with
Massey (Massey, Otto, & Antón, 2009), later
reviewed specifications of legally compatible
systems and produced requirements to foster legal
compatibility. However, those authors focused on
goals to specify requirements for the development of
legally compatible systems. Our approach to the
specification of such systems is broader in terms of
the views over legal texts it considers (the RSL
views).
Hoffmann, et al. (Hoffmann et al., 2012)
presented in 2012 some research on the
commonality within legal software requirements and
proposed legal software requirements patterns
(extracted from recurring legal requirements) to
produce catalogues of that kind of requirements. The
research reported in this paper involved dealing with
linguistic patterns and linguistic styles, not for legal
software requirements in the broad sense, but for the
specification of personal data protection
requirements, also extracted from legal documents to
produce a personal data protection requirements
catalogue.
6 CONCLUSION
The analysis of the GDPR is a starting point for the
further analysis, design and implementation of
information systems capable of ensuring the
confidentiality, integrity, availability and resilience
of the personal data they process. Due to
technological evolution, several services currently
share data and part of it relies on personal data
related to banking, healthcare and other data
domains. Those data require strict measures to
protect them from diversion for improper purposes
with irreversible consequences, yet maintaining the
free movement of such data. The application of
penalties for improper diversion purposes comes
from the detection of infringements, so the
extraction of business and system requirements from
the GDPR is of extreme importance, in order to
specify and further develop regulatory systems able
to communicate with processing systems and
operationalize the regulatory capabilities disposed in
the GDPR. This paper presented a systematic
approach to the analysis of the GDPR from both
business and system perspectives, useful for the
development and operation of regulatory systems.
The paper illustrated the approach with a sample of
some work products originated from executing the
analysis methodology.
Future work includes further analysis of the
GDPR from the system perspective, namely
involving the specification of Actors, DataEntities
and StateMachines views, as well as developing
automatic transformations of requirements into
formats other than the RSL Excel Template, along
with complementary diagrams. These model
transformations will require the use of RSL in its
programmatic shape, therefore transforming the
RSL-based SRS document in Excel format into an
even more rigorous representation. The RSL-based
specification of the GDPR will act as a catalogue of
personal data protection requirements and the
ultimate goal is for it to be reused in any project of
personal data regulatory systems development. In
future research efforts, the methodology to analyse
legal documents presented in this paper may be
applied to other regulatory documents that impose
requirements to information systems specification
and further development e.g. ISO 27000.
ICEIS 2018 - 20th International Conference on Enterprise Information Systems
404
ACKNOWLEDGEMENTS
This work was partially supported by national funds
under FCT projects UID/CEC/50021/2013 and
CMUP-EPB/TIC/0053/2013.
REFERENCES
Caramujo, J., & Silva, A. R. d. (2015). Analyzing Privacy
Policies Based on a Privacy-Aware Profile: The
Facebook and LinkedIn Case Studies. 17th IEEE
Conference on Business Informatics (CBI), Lisboa,
Portugal.
Regulation 2016/679 of the European Parliament and of
the Council of 27 April 2016 on the protection of
natural persons with regard to the processing of
personal data and on the free movement of such data
(General Data Protection Regulation) (2016).
Ferreira, D. d. A., & Silva, A. R. d. (2012). RSLingo: An
Information Extraction Approach toward Formal
Requirements Specifications. 2nd IEEE Model-Driven
Requirements Engineering Workshop (MoDRE),
Chicago, Illinois, USA.
Ferreira, D. d. A., & Silva, A. R. d. (2013). RSL-IL: An
Interlingua for Formally Documenting Requirements.
3rd International Workshop on Model-Driven
Requirements Engineering (MoDRE), Rio de Janeiro,
Brazil.
Hoffmann, A., Schulz, T., Hoffmann, H., Jandt, S.,
Roßnagel, A., & Leimeister, J. M. (2012). Towards the
Use of Software Requirement Patterns for Legal
Requirements. 2nd International Requirements
Engineering Efficiency Workshop (REEW), Essen,
Germany.
Massey, A. K., Otto, P. N., & Antón, A. I. (2009).
Prioritizing Legal Requirements. 2nd International
Workshop on Requirements Engineering and Law
(RELAW), Atlanta, Georgia, USA.
Otto, P. N., & Antón, A. I. (2007). Addressing Legal
Requirements in Requirements Engineering. 15th
IEEE International Requirements Engineering
Conference (RE), New Delhi, India.
Pohl, K. (2010). Requirements Engineering:
Fundamentals, Principles, and Techniques. Berlin
Heidelberg: Springer-Verlag.
Ribeiro, A., & Silva, A. R. d. (2014a). Evaluation of XIS-
Mobile, a Domain Specific Language for Mobile
Application Development. Journal of Software
Engineering and Applications, 7(11), 906-919.
Ribeiro, A., & Silva, A. R. d. (2014b). XIS-Mobile: A DSL
for Mobile Applications. 29th Annual ACM
Symposium on Applied Computing (SAC), Gyeongju,
Korea.
Savić, D., Vlajić, S., Lazarević, S., Antović, I., Stanojević,
V., Milić, M., & Silva, A. R. d. (2015). Use Case
Specification Using the SilabReq Domain Specific
Language. Computing and Informatics, 34(4), 877-
910.
Silva, A. R. d. (2015). Model-driven engineering: A
survey supported by the unified conceptual model.
Computer Languages, Systems & Structures,
43(October 2015), 139-155.
Silva, A. R. d. (2017). Linguistic Patterns and Linguistic
Styles for Requirements Specification (I): An
Application Case with the Rigorous RSL/Business-
Level Language. 22nd European Conference on
Pattern Languages of Programs (EuroPLoP), Irsee,
Germany.
Silva, A. R. d. (2018). Rigorous Requirements
Specification: Specification of Use Cases with the
RSLingo RSL Language. Lisboa: INESC-ID.
Silva, A. R. d., Caramujo, J., Monfared, S., Calado, P., &
Breaux, T. (2016). Improving the Specification and
Analysis of Privacy Policies: The RSLingo4Privacy
Approach. 18th International Conference on Enterprise
Information Systems (ICEIS), Rome, Italy.
Silva, A. R. d., Paiva, A., & Silva, V. (2018). Towards a
Test Specification Language for Information Systems:
Focus on Data Entity and State Machine Tests. 6th
International Conference on Model-Driven
Engineering and Software Development
(MODELSWARD), Funchal, Madeira, Portugal.
Silva, A. R. d., Saraiva, J., Silva, R., & Martins, C. (2007).
XIS-UML Profile for eXtreme Modeling Interactive
Systems. 4th International Workshop on Model-Based
Methodologies for Pervasive and Embedded Software
(MOMPES), Braga, Portugal.
Specification of Personal Data Protection Requirements
405