INFORMATION SYSTEMS SECURITY BASED ON BUSINESS
PROCESS MODELING
Joseph Barjis
Delft University of Technology, Jaffalaan 5, 2628 BX Delft, The Netherlands
Keywords: Secure business process modeling, Secure information system design, Information system security,
Organizational semiotics, DEMO methodology.
Abstract: In this paper, we propose a conceptual model and develop a method for secure business process modeling
towards information systems (IS) security. The emphasis of the proposed method is on social
characteristics of systems, which is furnished through association of each social actor to their authorities,
responsibilities and obligations. In turn, such an approach leads to secure information systems. The resulting
modeling approach is a multi-method for developing secure business process models (secure BPM), where
the DEMO transaction concept are used for business process modeling, and the Norm Analysis Method
(organizational semiotics) for incorporating security safeguards into the model.
1 INTRODUCTION
Traditionally, security measures have been
considered a technical issue to be dealt with when
the system architecture, hardware performance,
databases and software design are tackled
(Backhouse & Dhillon, 1996), and mainly addressed
in the implementation (coding, programming) phase
of the life cycle. But the fact is that security is more
of organizational issues and should be developed as
early in the IT application life cycle as in the
business modeling phase (Backes et al., 2003; Mana
et al., 2003; Herrman & Herrman, 2006). The
motivation for this is obvious. Information systems
are developed to enable certain business processes
within an enterprise context. Thus, security
measures should be applied while IS design is in its
early stages in the life cycle such as the process
modeling phase, where business analysts and
security experts have to identify security safeguards
and implement them in the business process models.
Business processes are the starting point for any
system development process (e.g., IS projects, IT
applications, software design). Practitioners of
software development project refer to a business
process as a software system blueprint (Nagaratnam
et al., 2005). As such, business processes are a
natural phase in the application development life
cycle, where security safeguards should conceive.
2 RELATED WORK
Information security and secure business process
modeling attracted many researchers. There have
been some research works done in this regard, where
authors address the need for sliding security aspects
down the system design ladder to business process
modeling phase (Mana et al., 2003; Backes et al.,
2003; Herrmann & Herrmann, 2006; Rodríguez et
al., 2007). This paper further advances the study of
security driven business process modeling by
introducing an innovative method and approach, and
proposes a conceptual model for secure BPM.
Incorporation of security safeguards in business
process modeling greatly increases the likelihood of
an adequate and secure system development.
Research and practical findings at IBM, reported in
(Nagaratnam et al., 2005), argue that business
process modeling is an ideal time in the life cycle of
software system development to begin capturing the
business security requirements that address any
security concerns that relate to the business.
The existing approaches to secure business
process modeling are predominantly based on the
extension of existing modeling notations and
methodologies with security tags and properties.
One of the earliest works tackling security issues of
information system using existing methodologies is
(Backhouse & Dhillon, 1996). In this work the
authors propose a concept for using the notions of
213
Barjis J. (2009).
INFORMATION SYSTEMS SECURITY BASED ON BUSINESS PROCESS MODELING.
In Proceedings of the 11th International Conference on Enterprise Information Systems - Information Systems Analysis and Specification, pages
213-218
DOI: 10.5220/0002006502130218
Copyright
c
SciTePress
authorities and responsibility in order to ensure that
activities are carried out by the authorized actors.
The approach and method used in (Firesmith,
2003) is based on the extension of UML. In
particular, the author proposes a method to derive
security use cases in order to model a problem
domain for secure application development.
Actually, due to its popularity as a requirements
elicitation method, the security use case approach
has been researched and developed by many
researchers.
Another popular and widely used modeling
language and method extended with security
properties is BPMN (business process modeling
notation). In (Rodríguez et al., 2007), the authors
integrate security requirements through business
process modeling. In particular, they propose a
BPMN extension to business process diagrams.
One of the dominant reasons that the role of
business process modeling in IS security is
undermined is because often the methods restrict
themselves to merely the conceptual and semantic
levels and, therefore, present little pragmatic value
for information system designers and developers. By
using the existing methods, it is difficult to
automatically analyze the models and, therefore, it is
not possible to test and simulate the embedded
security measures. To elevate the importance and
pragmatic value of security-driven business process
modeling, it is required that the models possess
certain qualities. First of all, the resultant model
should be amenable to test and simulation in order to
capture how and when security safeguards will be
triggered and enacted. Secondly, the models should
capture social roles, authorities and responsibilities
pertaining to each action. Thirdly, it is imperative
that the models capture interactions between
different entities (human actors, business units,
applications) to identify the level of security
sensitivity (e.g., access and modification of sensitive
data or inter-organizational transactions may be of
special security scrutiny). These qualities are the
research motivations and drivers for this paper. In
this paper, it is attempted to show that the proposed
method and approach for developing secure business
processes yield the mentioned qualities to a certain
extent and advances the existing experience from
both a theoretical and an application perspective.
The contribution of this paper is the proposal of a
conceptual model for developing secure business
processes, an approach to implement the conceptual
model, and a secure business process modeling
method. The advantage of the proposed method is its
underlying formal semantics, which allows models
to be automatically analyzed and simulated. In the
proposed approach, emphasis is made on the social
characteristics of the system by associating each
social actor to their authorities, responsibilities and
obligations. In this paper we use the DEMO
methodology transaction concept (Dietz, 2006) for
business process modeling, and the Norm Analysis
Method (Stamper, 1994) for incorporating security
safeguards into the model.
3 CONCEPTUAL MODEL
The proposed conceptual model for secure BPM,
illustrated in Figure 1, has two main components
that need to be developed and combined to create a
secure business process model. The first component
consists of the ‘business transactions’ that needs to
be identified based on a ‘business processes
description’. The second component is ‘security
safeguards’ that are mainly defined based on
security determiners (see below for definition) and
represents a set of security safeguards that are
defined in conjunction with each business
transaction. These two components are developed in
a collaborative manner (see Figure 2) and in
correlation with each other. Together, the two
components create a secure BPM, as depicted in
Figure 1 and enclosed into the dashed-line rectangle.
Figure 1: Conceptual model of security embedded BPM.
Security determiners – these are rules,
procedures, laws, and other measures that an
organization wants to be implemented with regard to
certain activities, processes, and roles.
For example: The ‘security determiners’ define if
a transaction execution involves any action on
sensitive personal records (read, delete, modify), or
whether a transaction is executed in the boundary of
two organizations requiring more security
(transmitting credit card information, health care
records), and so on. For each such transaction,
security rules, security precautions, and security
alerts are formulated at the concept level.
Once security safeguards are defined, business
transactions are coupled with their corresponding
ICEIS 2009 - International Conference on Enterprise Information Systems
214
security safeguards and every time a transaction is
initiated it will trigger the security rules to be
satisfied. Actually, security safeguards need not be
implemented for each business transaction. In fact,
analysts and designers are required to consider a
trade-off between the level of security robustness
and business processes performance.
There are two other components in the proposed
conceptual model. The ‘security logs’ component
contains records of all business transactions that
carry security-sensitivity and trigger security
safeguards. This component will allow monitoring
and managing access and execution of sensitive
transactions as well. The ‘business processes
description’ can be found in the existing
documentation or be prepared by the analysts where
the underlying business processes are described. An
example of such description will be presented later
when the FHCC case is discussed.
The approach we propose to implement the above
conceptual model is based on the collaboration of
different expertise that allows developing a secure
BPM in correlation with the components specified in
Figure 1. The proposition is that for secure BPM,
close collaboration of the three types of expertise
should be considered crucial. This approach is
illustrated through a secure BPM collaboration
triangle in Figure 2. As depicted, in the collaboration
triangle for secure BPM, the ‘business analyst’ role
is emphasized as prominent and central. Essentially,
for a successful secure BPM, it is important that
although the business modeling phase is led by
business analysts, it is carried out in collaboration
with security experts and domain users. Each of
these groups plays an essential role in developing
secure models: the ‘business analyst’ leads the
work; the ‘domain users’ provide domain knowledge
to the analysts and security experts; ‘security
experts’ support the business analysts identifying
and incorporating security safeguards into the
models. The outcome of these collaborative efforts
is identification of the security safeguards at a high
abstraction level. It is rather flagging security areas
for more detailed and robust analysis and
implementation as the analysis and design activities
unfold.
For example: In developing secure BPM for a
hospital, the business analyst (maybe also in
collaboration with a security expert) determines
which policies apply to a given activity in the
context of the business process. The business analyst
might model the requirement to control authorized
access to a business activity such as medical records
and to ensure that the medical information flow is
protected from unauthorized access and ensures
confidentiality. The medical security requirements
can be defined within the secure BPM. These
models rather provide a reference that may be used
by the hospital compliance officers, such as security
auditors, to verify and monitor adherence to the
hospital security and confidentiality policies.
Secure
BPM
Business Analyst
Figure 2: A secure BPM collaboration triangle.
Although the proposed conceptual model and
approach is fairly specific and detailed security
aspects are covered, this approach is still abstract
enough not to limit the developers and security
experts' freedom of implementation.
To summarize this section, it should be stated
that in our approach, business processes are
described in terms of business transactions, where
each transaction entails interaction of actors (human
actors), business units, and artifacts (IT
applications). Depending on the sensitivity of the
carried activities, transactions are complemented
with security safeguards. Since interaction of
components (actors, artifacts, agents) is the center-
point of our approach, the DEMO business
transaction concept (Dietz 2006) is used as a
theoretical basis for identifying business activities.
Next to the DEMO methodology, Petri nets are
adopted for constructing the model diagrams. In fact,
due to its formal semantics, models based on Petri
net notations are formal and lend themselves to
automatic analysis. Security safeguards are defined
using the Norm Analysis Method of Organizational
Semiotics (Stamper, 1994). The Norm Analysis
Method is adapted to define rules, norms,
authorities, responsibilities, and exceptions for the
execution of each transaction. Both the DEMO
Methodology and the Norm Analysis Method are
discussed in the following sections.
4 BUSINESS TRANSACTION
According to the DEMO Methodology (Dietz,
2006), a business transaction is a generic pattern of
action and interaction. An action is a productive act
INFORMATION SYSTEMS SECURITY BASED ON BUSINESS PROCESS MODELING
215
and represents an activity that brings about a new
result. An interaction is a communicative act
involving two actor roles to coordinate and negotiate
an action.
As depicted in Figure 3, the process in which a
transaction is completely carried out consists of
three phases:
- Order phase (O), during which an actor makes
a ‘request’ for a service or goods towards another
actor. According to DEMO, this phase represents a
number of communicative acts or interactions. This
phase ends with a commitment (‘promise’) made by
the second actor, who will deliver the requested
service or good.
- Execution phase (E), during which the second
actor fulfills its commitment, i.e., ‘produce’ the
service or goods. According to DEMO, this phase
represents a productive act.
- Result phase (R), during which the second actor
does ‘present’ the first actor with the service or
goods prepared. According to DEMO, this phase
also represents a number of communicative acts or
interactions. This phase ends with the ‘accept’ of the
service or goods by the first actor.
In Figure 3, by arrows and circles, it is also
illustrated that all ‘request’, ‘promise’, ‘present’,
‘accept’ acts can be logged in the model.
As it becomes obvious from the three phases,
each business transaction is carried out by two actor
roles. The actor role that initiates a transaction is
called the ‘initiator’ (e.g., customer) and the actor
role that executes the transaction is referred to as the
‘executor’ (e.g., supplier) of the transaction.
Executor
Execution
phase
Initiator
Result phase
Order phase
Request
Produce
Present
Promise
Accept
Figure 3: Business transaction pattern.
4.1 Illustrative Example
It should be noted that the case study discussed here
is a simplified description of the processes in which
some details are omitted and skipped. This
description is based on a family health care center
(FHCC), taken from (Barjis & Hall, 2007).
In order to be examined by a doctor, a patient
needs to make an appointment beforehand. On the
appointment day, a nurse first conducts preliminary
general checkup (blood pressure, EKG, basic lab
work) and ‘records’
chief complaints, and reason
for the visit. After completing this preliminary exam,
the nurse escorts the patient to an available
examination room. The doctor examines the patient
and ‘updates’
the patients chart if any prescription
is issued, diagnosis is made, referral is given, or if
any other notes are taken. After completing the
examination, the patient goes to the side-desk to
check out, to make the (co-)payment relevant to the
service delivered. In rare cases, patients may need
further examination by external healthcare
providers (sub-specialist) or advanced diagnostic
equipment such as a CAT scan, available elsewhere.
In this case, the FHCC schedules an appointment
with the external healthcare provider based on the
availability of the network provider. Some
procedures such as a CAT scan may require the
insurance company’s pre-approval.
4.2 FHCC Business Transactions
Making an appointment is the first activity in the
series of processes taking place in the “patient
examination.” By making an appointment, a new
fact is created, being that a new appointment is made
and recorded into the system. In this activity, the
patient is the initiator and the receptionist is the
executor. This activity comprises the first business
transaction (T1) in the process of “patient
examination.” Thus, this and the other main
activities are described as follows:
T1:
Initiator:
Executor:
Fact:
making an appointment
patient
FHCC (receptionist)
a new appointment is made
T2:
Initiator:
Executor:
Fact:
requesting health care
patient
FHCC (physician)
patient is given health care
Transaction 2 requires that the patient records
will be accessed and modified by a physician.
Therefore this transaction requires explicit access
authorization to secure the patient's electronic
records.
T3:
Initiator:
Executor:
Fact:
conducting general physical test
FHCC (physician)
FHCC (nurse)
general physical test is conducted
Similarly, Transaction 3 requires that the patient
records will be accessed and new records will be
added by a nurse. Therefore this transaction requires
explicit access authorization to secure the patients
electronic records.
ICEIS 2009 - International Conference on Enterprise Information Systems
216
T4:
Initiator:
Executor:
Fact:
arranging an external appointment t
FHCC
Specialist (external provider)
an external appointment is made
T5:
Initiator:
Executor:
Fact:
requesting a pre-approval
Specialist
Insurance
a pre-approval is granted
T6:
Initiator:
Executor:
Fact:
paying the bill
FHCC
patient
the service is paid
These business transactions constitute a network
of actions and interactions through which the FHCC
business processes emerge. Due to limited space, we
skip the FHCC patient examination process. Instead,
we would like to discuss a few points that relate this
model to the research questions and objectives we
identified in the beginning of the paper.
The resulting model using the transaction schema
presented earlier can be automatically analyzed and
simulated. However, the model would not include
any security safeguards. Therefore, the DEMO
transaction concept, as we discussed in the
conceptual model, needs to be complemented by
security safeguards, for which purpose we use the
NAM of the organizational semiotics. To do so, we
first introduce the NAM in the next section and then
revisit the DEMO transaction diagram for security
improvement.
5 NORM ANALYSIS METHOD
Organizational Semiotics is a framework and a set
of methods based on the understanding of
organizations as systems of social norms. It
emphasizes the central role of the people, their
responsibility and the organization in the analysis
and design of information systems (IS) (Stamper,
1994). Organizational semiotics consists of a set of
methods including the Norm Analysis Method
discussed in this section.
In the previous section, we illustrated how
business transactions can be identified and the
relevant actors defined. The Norm Analysis Method
is used as a suitable complement to deal with
behavioral norms specification. According to the
organizational semiotics framework, there are five
types of norm that influence certain aspects of
human behavior. They are ‘perceptual norms’,
‘cognitive norms’, ‘evaluative norms’, ‘behavioral
norms’ and ‘denotative norms’ (Stamper, 1994).
In business, most rules and regulations fall into
the category of behavioral norms. These norms
prescribe what people must, may, and must not do,
which are equivalent to three deontic operators “is
obliged,” “is permitted,” and “is prohibited.”
Therefore, particular attention is given to the
behavioral norms since they are expressed as
business rules, and have direct impacts on business
operations. Behavioral norms govern human
behavior within regular patterns. The following
format is considered suitable for specifying of
behavioral norms – a generic structure for all norms:
whenever
if
then
is
to
<condition>
<state>
<actor>
<deontic operator>
<action>
In this structure, the condition describes the
situation where the norm is to be applied, and
sometimes further specified with a state-clause (this
clause is optional). The actor-clause specifies the
actor responsible for the action. The actor can be a
staff member, a customer, or a computer system if
the right of decision-making was delegated to it. As
for the next clause, it is a deontic state and is usually
expressed by one of the three operators - permitted,
forbidden and obliged. For the next clause, it defines
the consequence of the norm. The consequence
possibly leads to an action or to the generation of
information for others to act.
Now using the above format along with
identified business transactions, we develop a secure
model of business processes. We will use capital “S”
for each security safeguard, where the number
following it indicates the number of the
corresponding transaction (e.g., S2 is definition of
security safeguards associated with Transaction 2).
Here, using the norms, we will discuss security
safeguards for only two business transactions: one
for the physician role that requires ‘update’
of the
patient medical records; one for the nurse role that
requires ‘record’
of new data for each visit by the
patient. Actually, in the same manner we can define
security safeguards for all roles and transactions, but
we will limit ourselves to only two transactions, in
part due to space limitations.
S2:
whenever
if
then
is
to
<a physician examines a patient >
<the physician is the patient family
doctor>
<the physician>
<permitted>
<update the patient medical records>
INFORMATION SYSTEMS SECURITY BASED ON BUSINESS PROCESS MODELING
217
S3a:
whenever
if
then
is
to
<a nurse records into the patient files>
<the nurse is authorized so>
<the nurse>
<obliged>
<sign the records with her e-signature>
S3b:
whenever
if
then
is
to
<a nurse updates existing records>
<the record is created in the past >
<the nurse>
<is prohibited>
<make changes to existing records>
All the safeguards can be diagrammatically
incorporated into the complete model of the FHCC
patient examination process. However, let us revisit
the business transaction diagram (Figure 3) and
show incorporation of security safeguards in more
generic terms and how and where these security
safeguards will be executed in a business
transaction.
Executor
E-phase
Initiator
Result phase
Order phase
Request
Produce
Present
Promise
Accept
Security
Rules
Figure 4: Business transaction with embedded security.
Figure 4 represents a generic form of business
transaction with an added feature modeled as
‘secure’. The secure box is basically implementation
of the security safeguards (i.e., the aforementioned
rules) defined for each transaction. As seen from the
diagram, it will be triggered before an actor is
allowed to perform or execute any action. In the
FHCC case, such a box will be implementation of
S2 before Transaction 2 is executed or
implementation of S3a and S3b before Transaction 3
is executed.
6 CONCLUSIONS
In this paper we have discussed security driven
business process modeling method that allows
constructing models based on the DEMO transaction
concept using formal graphical notations of Petri
nets. The security safeguards embedded in the model
are developed using the Norm Analysis Method of
organizational semiotics. Although due to space
constraints, the complete business process model of
FHCC and simulation are skipped in this paper, the
resultant models can be simulated in order to
observe how the security safeguards are called and
executed to ensure authorized access before security
sensitive actions are executed. For example, if a
nurse wants to modify existing medical records, this
action will trigger a security measure that the nurse
should be able to comply with. This security
measure can be another level of access. An
advantage of the proposed method is that the models
can be simulated and dynamically tested in regard to
security safeguards. As each secure transaction is
executed, the corresponding security safeguard is
deployed generating security logs for future analysis.
REFERENCES
Backes, M., Pfitzmann, B., & Waidner, M. (2003).
Security in Business Process Engineering. In
Proceedings of 2003 International Conference on
Business Process Management. Lecture Notes in
Computer Science vol. 2678, Springer.
Backhouse, J., & Dhillon, G. (1996). Structures of
responsibility and security of information systems.
European Journal of Information Systems, 5, 2–9.
Barjis, J., & Hall, M. (2007). A Healthcare Center
Simulation Using Arena. In the proceedings of
MSVVEIS’07”, June 12-13, Funchal, Madeira –
Portugal.
Dietz, J.L.G. (2006). Enterprise Ontology –Theory and
Methodology. Springer.
Firesmith, D. (2003). Security Use Case. Journal of Object
Technology, Vol. 2 (3), pp. 53-64.
Herrmann, P., & Herrmann, G. (2006). Security
requirement analysis of business processes. Electronic
Commerce Research, Vol. 6 (3-4), pp. 305-335.
Mana, A., Montenegro, J.A., Rudolph, C., & Vivas, J.L.
(2003). A business process-driven approach to security
engineering. Proceedings of the 14th International
Workshop on Database and Expert Systems
Applications, pp. 477-481, Prague.
Nagaratnam, N., Nadalin, A., Hondo, M., McIntosh, M., &
Austel, P. (2005). Business-driven application
security: From modeling to managing secure
applications. IBM Systems Journal, Vol. 44, No 4.
Rodríguez, A., Fernández-Medina, E., Piattini, M. (2007).
A BPMN Extension for the Modeling of Security
Requirements in Business Processes. IEICE -
Transactions on Information and Systems, Volume
E90-D, Issue 4, Pages: 745-752.
Stamper, R. K. (1994). Social Norms in Requirement
Analysis – an outline of MEASUR. In Jirotka, M., &
Gorguen, J. (Eds.) Requirements Engineering: Social
and Technical Issues.
ICEIS 2009 - International Conference on Enterprise Information Systems
218