ABSTRACT: This paper proposes a viewpoints framework for
deriving organisational information requirements in a manner that
takes into account the need to continuously review and capture
information needs as they change over time. The viewpoints are
endogenous forces emanating from within a balancing-loop of an
organisational information system that cause dynamics in its information
requirements. The viewpoints captured are used as the basis for
qualitative modelling and performing quantitative assessment of
the system's functional and non-functional performance. System
Dynamics has developed a role as a methodology for modelling the
behaviour of complex socio-economic systems. The viewpoints conceptual
framework proposed in this paper provides a powerful approach
to understand and improve the requirements engineering process
of an organisational information systems as a useful way to improve
its effectiveness.
Key words: Conceptual Modelling, Organisational Information
Systems, Requirements Engineering Process, System Dynamics, Viewpoints.
1. INTRODUCTION
Organisational information systems (OIS) have become an important
area of research. Conceptual models of OIS have been applied to
many aspects of the information systems (IS) development life
cycle, particularly addressing issues of requirements engineering
(RE) (Wieringa, 1995). Loucopoulos and Karakostas (1995) define
conceptual models as cognitive structures used for purposes of
understanding and communicating aspects of the physical and social
world around us. The above definition forms a basis for the viewpoints
conceptual framework for capturing and improving the RE process
in OIS.
The RE process has been the focus of many recent organisational
systems' projects (Macaulay, 1996). These changes in the organisations'
environment affect organisational performance and indeed OIS effectiveness
over time. The two main problem areas in RE process are problem
analysis and system description (Davis, 1993). While the first
issue is characterised by uncertainty, an expansion of information
and knowledge, the second is characterised by organisation of
ideas, resolving conflicting views and eliminating inconsistencies
and ambiguities (Macaulay, 1996).
Systems dynamics (SD) has become an important methodology for
understanding and formalising conceptual process models (Abdel-Hamid
and Madnick, 1990). The conceptual framework proposed by the authors
recognises the role of systems thinking and the application of
SD (Forrester, 1961; Meadows, 1982; Coyle, 1986; Senge, 1993;
HPS, 1994) within the traditional requirements engineering process.
In this paper, viewpoints are defined as expressed organisational
information needs of strategic, tactical and operational management
levels within an OIS necessary to achieve organisational business
objectives. It is long established that most information systems
fail to meet the performance expectations of users at different
levels of management (Munro and Gordon, 1977; Anthony, 1977).
This highlights the inadequacies of traditional requirements analysis
techniques coping with complex and dynamic changes in user viewpoints
over the OIS life cycle.
2. ORGANISATIONAL STRUCTURE AND MANAGEMENT ACTIVITY
RE process in OIS must embrace methods and techniques that support
understanding of organisations and management information needs.
Organisation and management literature highlight the fact that
information needs are different at various levels of management
(Anthony, 1977), and the nature of decisions is different too
(Ackoff, 1981). Anthony (1977) further shows how failure to make
distinctions between information needs for strategic planning,
management control and operational control, have led to mistakes
in designing and using systems. Mintzberg's (1979) synthesis of
the research on organisational design is very important in understanding
of organisational structures and their processes.
Current methods used in OIS can be designed to produce different
modes of information. However, it is difficult to use these traditional
methods to design a system that takes into account information
needs for all its users. Although several methods and techniques
have been proposed to support requirements engineering process
(Bubenko et al, 1994; Yadav, 1983; Lyytenin, 1981; Mumford, 1984;
Wieringa, 1996; Macaulay, 1996), most of these are suitable for
specifying design requirements. They are not effective for eliciting
information requirements of an OIS. The use of systems thinking
and the application of SD modelling notation (Abdel-Hamid and
Madnick, 1990; Meadows, 1992; Peterson, 1992) to describe changes
in organisational structures and their processes, may be useful
in eliciting and capturing the information requirements an OIS.
3. USE OF SYSTEM DYNAMICS
SD has developed a role as a method for modelling the behaviour
of complex socio-economic systems (Forrester, 1961; Keys, 1988;
Coyle, 1995). SD notation can be applied to build detailed conceptual
model of organisational processes (Meadows, 1982; Pidd, 1992)
and therefore identification of information needs at different
levels of managerial activity. The viewpoints are the endogenous
forces emanating from a system's structure (Forrester, 1987) within
a balancing-loop (HPS, 1994) of an organisational information
system's that causes the dynamics in its information requirements.
The use of systems thinking and application of SD modelling notation
in RE process provides the analysts with a balanced perspective
between "hard" (Davis and Vick, 1977; Boehm, 1981) and
"soft" (Checkland and Scholes, 1990) systems problem-solving
paradigms. The hard systems approach (e.g Structures Systems Analysis
and Design method and Operational research) are too detailed and
support a narrowly focused operational viewpoint. "Hard"
systems thinking approach is unable to tackle adequately problems
which are ill structured (Keys, 1988). While the soft systems
approach (e.g Soft Systems Methodology) is a very broad systems
viewpoint and supports a cosmic perspective. The "soft"
systems approach sees the real-world situations as inherently
complex and difficult to understand so that it cannot permit effective
operating contact with reality. The balanced perspective provided
by SD to handle both "hard" and "soft" systems-based
problems (Meadows, 1982; Wolstenholme, 1992), seems to break the
paradigm incommensurability (Kuhn, 1970). The SD conceptual model
is a "bridge" between the non-technical world view of
the user and the technical world view of designers and programmers.
4. VIEWPOINTS CONCEPTUAL FRAMEWORK
This framework uses influence diagram-based technique to assist
with the process of requirements elicitation. Influence diagrams
provide a systematic approach to capturing data and structure
(Senge, 1993) and identifying knowledge gaps.
The viewpoints framework combines traditional business and systems
analysis techniques and high level SD conceptual models of the
OIS performance over its life cycle. As individual system viewpoint
changes, information requirements also change and these changes
are compared to organisational performance objectives. The use
of information feedback systems facilitates OIS performance measurement.
As the performance gap in its effectiveness increases, so is the
need to review and change requirements. The process is continuously
monitored and controlled, and improvements in the OIS performance
introduced as the new business rules are defined and specified
in a properly managed RE process.
The proposed framework can be explained in terms of operational
thinking (Senge,1993; HPS, 1994) using influence diagram. Due
to mounting pressure to adjust desired performance, the
management interacts with external organisational factors
to produce changes in business objectives. These changes
after some time influence changes in the business rules.
Changes in organisational business rules influence user
viewpoints to increase pressure on management to adjust
desired performance after a delay period. Performance indicators
introduced in desired system performance interacts with
current or planned performance to reduce a performance
gap. The larger the gap the stronger the influence to introduce
RE process.
The RE process influences the move of current or planned
performance and, after some delay, user satisfaction.
Time delay affects the user satisfaction which subsequently
impedes the migration of the current or planned performance
in the intended direction. As current or planned performance
moves to desired performance the performance gap
is reduced towards zero. The desired system performance
is undermined by the current or planed performance as it
works to reduce the performance gap lessening the influence
towards RE process. Finally, the system reaches an equilibrium
(haemostasis) other than what was the initial desired performance.
Due to defined organisational structures and their processes in
an OIS, responses to patterns of behaviour in their environments
ensure that the system sends control messages to their environment.
Messages are sent either to analysts or other control meta-systems,
responsible for RE process. In return, information requirements
document is updated through information feedback mechanisms identified
in the framework (Figure 1).
5. SUMMARY AND CONCLUSION
This paper proposes a conceptual framework for identifying and
improving organisational information requirements in the context
of system dynamics. The concept of continuous review of IR is
emphasised as IR changes over time in the OIS life cycle. The
fundamental understanding that emerges from this framework is
that the effectiveness of OIS is affected by the organisational
context to which they are applied (Mintzberg, 1979; Munro and
Gordon, 1977; Macaulay, 1996; Loucopoulos and Karakostas, 1996).
To improve the effectiveness of OIS, the framework identifies
that:
RE in OIS cannot be considered apart from organisational context
(Yadav, 1983). IR change over time as both the organisation and
its environment are dynamic in nature. Viewpoints of all system
users and stakeholders must be taken into account for the resulting
OIS to be effective. Of course taking into account the conflicting
interests of stakeholders and its cost effectiveness. Deriving
OIR must take into account the complexity and dynamic nature of
viewpoints and their changes over time in the OIS life cycle.
The use of system dynamics high level models to handle both hard
(Boehm, 1981; Pidd, 1982) and soft organisational data (Checkland
and Scholes, 1991; Senge, 1993) and its ability to facilitate
understanding and learning, is an effective way to improve RE
process of organisational information systems
The authors contend that this proposed conceptual model provides a powerful framework for improving the requirements engineering process, particularly for information requirements of OIS. Empirical studies will be vital in testing the usefulness of system dynamics in improving the requirements engineering process. Results of such studies and further refinement of the viewpoints framework will be reported in subsequent papers.
REFERENCES
[For the list of references please see the virtual proceedings]
(http://www.scism.sbu.ac.uk/~williadw/viewponts.html