Shuyan Xie
School of Computing, Dublin City University,Glasnevin, Dublin 9, Dublin, Ireland
Markus Helfert
School of Computing, Dublin City University,Glasnevin, Dublin 9, Dublin, Ireland
Keywords: Information Architecture, Information Quality, Inter-organizational Service, Emergency Medical Service.
Abstract: In inter-organizational service (IOS) system, the quality of information exchanged and shared by involved
organizations is important. Although information quality (IQ) has been emphasized for decades, IQ
problems still widely exist. For a significant class of information related to semantic issues, it is necessary to
improve information quality not just by working on the information/data itself. However, this is not
commonly understood and often leaves little doubt about the effectiveness of the current approach. We
consequently propose an architectural approach to enhance IQ for IOS: enterprise-level information
architecture allows a rich contextual environment to guarantee IQ, and provides a traceable path to measure
IQ across organizations. It is demonstrated in an emergency medical service enterprise.
In the class of applications that heavily depends on
information quality (IQ), a typical approach to solve
the IQ problem usually starts and ends with the
activities scoped to the physical data storage level
(Wang et al., 1995). But most of the efforts have
only been successful to a certain degree (Drake et
al., 2004). Given that in business applications
information exists within the context of business
processes, the attempts to solve IQ problems at the
purely physical information level are not effective
(Drake et al., 2004). The physical level does not
capture the requisite semantics to accurately
communicate information across processes. As a
result, most of the semantic information issues exist
at the exchange processes and organizational
boundaries. For inter-organizational service (IOS)—
we see it embedded in one enterprise—the top
(enterprise) level is the focal point with the highest
probability for data discrepancy (Eden and Kazman,
2003). Enterprise level information models are
practically absent in current organizations, and
therefore lacks of effective communication for the
information across enterprise wide, organization or
service boundaries.
This paper proposes to extend the information
centric approach and focuses on creating an
enterprise architectural view to analyze information.
The reminder of the paper is structured as follows:
Section 2 outlines the research methodology.
Background is discussed in section 3. Section 4
presents our information architecture model, which
is embedded in an Emergency medical service
(EMS) case. We conclude our paper in section 5 by
summarizing the main contributions together with
some remarks for further research.
This research is mainly concerned with theory
building and thus it can be classified as being
interpretive in nature. Interpretative research does
not predefine dependent and independent variables,
but focuses on the full complexity of human sense
making as the situation emerges (Anderson et al.,
2005). The architectural approach for IQ analysis
Xie S. and Helfert M..
DOI: 10.5220/0003503604380443
In Proceedings of the 13th International Conference on Enterprise Information Systems (ICEIS-2011), pages 438-443
ISBN: 978-989-8425-56-0
2011 SCITEPRESS (Science and Technology Publications, Lda.)
presented in this paper has been illustrated by
reviewing existing relevant literature in the domain
of data, information quality, architecture, enterprise
architecture (EA) and IA. The theoretical proposal
presented in this paper consists of concepts and their
relationships, as identified from literature, which
have been further demonstrated in real case study.
Thus, this research can be described as being
interpretive and grounded in literature.
3.1 Information Quality
Information is difficult to manage from business
standpoint. More specifically, it is not the
information itself that is difficult, the problems lie
with the people and processes using that
information. So, in order to improve IQ under
sharing environment across organizations, we must
focus from dynamic perspective: the process of
distributing information. Common definition of IQ is
“fitness for use”. Information are high quality if they
are fit for their intended uses in operations, decision
making, and planning (Wang, 1998). In this sense,
when information shared and exchanged among the
parties are deemed of high quality if free of defects
and process desired features. It point to the notion
that IQ needs to be measurable and being measured
appropriately while they are shared by multiple
Data is produced by measurements or
observation (Drake et al., 2004), which brings to an
important concept—a notion of data context or
metadata—that is critical to the success of IQ
improvement efforts. One of the causes for poor IQ
problems is the lack of sufficient information
context. To solve the poor IQ problem, information
context should be defined and well understood
(Drake et al., 2004, Eden and Kazman, 2003).
3.2 Architecture Approach
The word architecture is used whenever a high-level
overview of interrelated components wanted to be
defined, and when the relationships among them are
complex and difficult to understand. Architecture is
generally can be captured as a set of abstractions
about the system that provides enough essential
information to form base for communication,
analysis, and decision making. (Foegen and
Battenfeld, 2001, IEEE, 2000, Kazman et al., 1996,
Kruchten et al., 2006, Rechtin, 1992). This points to
the notion that architecture allows better
understanding on the components and their
relationships. In-depth discussion of various
architectures is out of scope for this paper. We focus
on EA to analyze enterprise-level IA.
Various EA frameworks define several views
that focus on specific aspects such as business,
technology, information, and so on, to reduce the
complexity. IA has been indicated as one important
component for EA (IEEE, 2000, Laudon and
Laudon, 2002). IA defines and establishes the
information component of the EA by providing
abstract representations of corporate information.
This is where information requirements are specified
at a high level, typically as subject areas, entities,
and relationships. In doing so, all other EA
components must be included. These relationships
characterize how and by whom data is used and
where it flows. The IA is used for understanding the
information needed and used by people in
performing tasks and business processes.
Information is created by processes and tasks and is
shared with other processes and tasks (Rood, 1994).
Figure 1 presents an understanding on the trend
and focus of IA based on the timeline. IA is
originated with static structure for information
management. Researchers initially identify and
employ the need for flexible IA in considering of the
dynamic information environment. With the
increasing dynamic requirements, researchers begin
to focus on IA framework to demands under
different situations (Campbell and Hummel, 1998,
Duncan and Holliday, 2008, Ray et al., 2003, Riva
and Rodriguez, 2002). From the time trend, it shows
increasing attention on dynamic aspects in terms of
the changeable and situational environment, as for
example the structure or design of the environment
with the data collection, data exchange. As dynamic
concept is developing, such concept is applied back
to and strengthened the context of static IA
(Sherman, 2002).
Definition of architecture has been discussed by the
researchers and practitioners, but there are no single
defined concept is accepted. In general, there are
two basic approaches that can be noticed regarding
these definitions, one sees architecture as a
descriptive concept (show as close circle in Figure
1) that factually describes the characteristics of
existing artifacts, whereas the other sees architecture
as prescriptive concept (show as open circle in
Figure 1) that defines how artifact should be
realized. It indicates that the former approach allows
information elements exchanged and shared to be
described and mapped for IQ assessment and
measurement; the later approach indicates that from
design point of view, enterprise-level IA allows a
contextual environment to design to guard the IQ.
Figure 1: Information Architecture Research.
From the discussion and analysis, we describe
that enterprise-level IA as a set of different
information elements so connected or related as to
perform a unique function, which is not performable
by the elements alone. All indicate that the
enterprise-level IA contains data definitions of the
enterprise constituencies as well as the relationships
among these constituencies. This again point to the
notion that architecture provides a consistent
contextual environment. We define that IA
represents/defines the structure of information,
including static aspects as a mapping showing the
information elements, the interfaces and
relationships between the various information
elements (IEs), and dynamic aspect that the
relationship of how the information shared and used
across organizations. Thus, it allows a path for better
IQ measurement by tracing the process.
4.1 Architecture as Metadata Source
Architecture can be seen as metadata source. A rich
contextual environment (metadata) needs to exist,
and a comprehensive set of models is needed to
produce these models. The EA modeling efforts can
produce these models (Fuller and Morgan, 2006).
Time-critical services such as EMS introduce
complexities to multi-organizational information
sharing, including the need for timely information in
a form that can be trusted and used by emergency
responders (Dawes and Prefontaine, 2003, Horan
and Schooley, 2007). IQ problems exists in EMS
(Fisher and Kingma, 2001), however they often find
it difficult to assess their current IQ. As mentioned
previously, a notion of information context is
absolutely critical to the success of any IQ
improvement effort. This notion is at the crux of the
poor information quality problem—insufficient
information context (metadata).
A comprehensive set of models is necessary to
produce this desired metadata. To work with such
architectural complexity, some decomposition
method s needed. One such method is the layered
model. The information layer is where elements
concerning information and data are captured and
managed. We introduce the layered model and give
an overview of the main constructs available for
modeling IA. A possible layering can constitute a
conceptual layer at the top, the logical layer in the
middle, and the physical layer on the bottom (Bruel
et al., 2002). This model assumes an information-
centered approach. Adapted from National
Intelligent Transportation System (ITS) Architecture
“provides a common structure for the design of
intelligent transportation systems” and prescribes a
general model that supports the development of
many different designs (2003). A simplified three-
layered model is shown in figure 2. As with the
other layers of the core meta model, it is split into
the following views:
Conceptual—where defines the ‘what’.
Information terms this means ‘what’ information
concepts are required within each domain.
Logical— where defines the ‘how’. In
information terms this is the next level of abstraction
down, where defines ‘how’ the information concept
are used. In this layer, it presents a functional view
ICEIS 2011 - 13th International Conference on Enterprise Information Systems
Figure 2: Enterprise-level Information Architecture Three-layered Model.
that consists of specifications that are used to
perform user services. In this view, the functions are
represented in a set of data flow diagrams.
Physical – The physical information view
captures a particular view of information managed.
It is a physical representation of important interfaces
and components, and it divides the logical
architecture functions into a number of high level
The model shows that enterprise information
models lie in both the conceptual and in the logical
layers, and provide the foundation for consistent
interaction between these layers.
Reflection to the National ITS Architecture
(2003), this IA framework is developed along two
dimensions: horizontal and vertical. Traceability
along both dimensions is necessary for the vertical
(between conceptual, logical, and physical) and
horizontally within each layer but across the
organizational boundaries. It is important to
emphasize the business process is needed to be
involved as the foundation for this IA model.
In the proposed three layered view of the IA, the
conceptual layer includes the highest possible level
of abstraction and therefore it captures the
foundational components and their relationships.
Thus, the top layer model is very stable and is not
subject to change unless the most essential
underlying structures change. The information
elements that are defined at this level are cross
referenced among the layers. Every element has at
least one conceptual definition that references it. The
reverse is also true: there is no information element
presented that does not exist in the conceptual layer.
In the EMS case, where EMS is seen as a single
enterprise that consists of multiple organizations,
each individual organization will need to have its
own three-layered organization-level IA model,
where top-level organizational information concepts
and the corresponding information elements will be
mapped unambiguously to the enterprise-level
model concepts. In this sense, only the elements that
have their counterparts at the enterprise level can be
possibly mapped. By relating each organizational
level definition to the common enterprise level
equivalents, we are eliminating semantic mismatch
between different organizations. Since the
information elements are cross-referenced with the
process specification, there is enough contextual
information to correlate information elements at the
enterprise and individual organizational levels.
The logical layer of this enterprise focused IA
model emphasizes information-related
considerations and defines specifications for
enterprise-level information. There are
data/information that needs to be constructed to
support the business processes defined at the top
layer of the model. By defining information
requirements in terms of the business processes,
another major cause of low data quality is
eliminated: the disconnect between the business and
information view (Mukhopadhyay et al., 1995).
Under EMS case which is also known as a multi-
organizational enterprise, it is quite common for
more than one system to be operating on information
elements from objects defined at the top conceptual
layer. Under this model, each system specification
will define its own unique information attribute, but
all these attributes are in turn mapped to the one
element at the top layer. This top-down
decomposition helps to alleviate a problem that is
similar to “departmental information silo”
(McGuffog, 1997).
Such proposed model for IA is necessary for IQ
analysis and improvement, especially for a complex
socio-technical enterprise like EMS enterprise that
involves a strong dynamic aspect of the information
4.2 Architecture Provides Dynamic IQ
In our description of information architecture, IA
indicates: (1) Information elements (IEs), (2)
Structure of IEs in an enterprise, (3) Information
relationship/flow/exchange among all the involved
organizations. We envision the structural
architecture from static and dynamic aspects for
EMS case. As showed in Figure 3, IEs (incident
information, patient information etc.) can be
presented as one way sequence, end-to-end
sequence, or two-way sequence while shared across
Organizations (components). By mapping the
structure in a static view and the information flow
process, it enables the dynamic consideration to
monitor information/data that is generally isolated
within each of the individual organizational
environment. It allows to detect the information
quality gap (accuracy, relevance, completeness etc.).
Figure 3 shows a high-level overview information
transmitted across organizations in emergency case.
As knowing the static information structure and the
dynamic information flow, we can systematically
analyze and measure the IEs in each process. The
way how information evolves and is connected
provides a path to trace complex information
relationship. We can trace the changes within and
across the data stores—Computer Aided Dispatch
(CAD), Patient care records (PCR), and Hospital
Information system (HIS)—that allows us to
measure the data quality from end-to-end, following
the concept of information manufacturing systems
that produces information products of which quality
can be measured (Pham Thi and Helfert, 2007).
The main contribution of this paper is proposing an
approach to describing enterprise-level IA and using
these descriptions to indicate that IQ can be
improved if architectural concept is enhanced.
Enterprise-level IA can be rich metadata source to
guard IQ as the information definition is traceable
across organizations vertically and horizontally;
Descriptive IA mapping allows IQ assessing and
measuring both statically and dynamically. A robust
traceability mechanism is necessary for high-quality
information. The architectural models provide a
foundation for the information traceability and thus
quality information. We demonstrate it within the
EMS case. This paper further confirmed the project
idea of that information focused architecture
provides a tool for information assessing and
measuring, and therefore improve the quality.
Figure 3: Information Flow across Organizations.
ICEIS 2011 - 13th International Conference on Enterprise Information Systems
2003. National ITS Architecture Executive Summary.
Federal Highway Administration, United States
Department of Transportation. Washington, D.C.,
Allen, B. & Boynton, A. 2008. Information Architecture:
in Search of Efficient Flexibility. MIS Quarterly, 15
(4), pp.435-445.
Anderson, J., Aydin, C., Kaplan, B. & Maxwell, J. 2005.
Qualitative Research Methods for Evaluating
Computer Information Systems. Evaluating the
Organizational Impact of Healthcare Information
Systems. Springer New York.
Bruel, J. M., Bellahsene, Z., Mellor, S., Scott, K., Uhl, A.
& Weise, D. 2002. Model-Driven Architecture.
Advances in Object-Oriented Information Systems.
Springer Berlin / Heidelberg.
Campbell, A. P. & Hummel, J. R. 1998. The dynamic
information architecture system: an advanced
simulation framework for military and civilian
applications. [Online] Available from
[Assessed 10 January].
Dawes, S. S. & Prefontaine, L. 2003. Understanding New
Models of Collaboration for Delivery Government
Services. Communications of the ACM, 46 (1), pp. 40-
Drake, D. B., Steckler, N. A. & Koch, M. J. 2004.
Information Sharing in and Across Government
Agencies. Social Science Computer Review, 22(1), pp.
Drucker, P. 1985. Management, New York, Haper.
Duncan, J. & Holliday, W. 2008. The Role of Information
Architecture in Designing a Third-Generation Library
Web Site. College and Research Libraries, 69(4), pp.
Eden, A. H. & Kazman, R. 2003. Architecture, Design,
Implementation. Proceedings of the 25th International
Conference on Software Engineering.
Evernden, R. & Evernden, E. 2003. Third-Generation
Information Architecture. Communication of the ACM,
46(3), pp. 95-98.
Fisher, C. W. & Kingma, B. R. 2001. Criticality of Data
Quality as Exemplified in Two Disasters. Information
& Management, 39(2), pp. 109-116.
Fuller, T. & Morgan, S. 2006. Data Replication as an
Enterprise SOA Anti-Pattern. The Architecture
Journal 8(16).
Garrett, J. J. 2002. Information Architecture. Digital Web
Horan, T. A. & Schooley, B. L. 2007. Time-critical
Information Services. Communication of ACM, 50, pp.
IEEE 2000. Standard for Architectural description.
Institute for Enterprise. IEEE 1471-2000. IEEE.
James, C. B., Larry, S. & Salvatore, T. M. 1989. Building
and implementing an information architecture.
SIGMIS Database, 20(2), pp. 9-17.
Kruchten, P., Obbink, H. & Stafford, J. 2006. The Past,
Present, and Future for Software Architecture.
Software, IEEE, 23, pp. 22-30.
Laudon, K. C. & Laudon, J. P. 2002. Management
Information Systems: Managing the Digital Firm,
New Jersey, Prentice Hall.
Mcguffog, T. 1997. The Obligation to Keep Value Chain
Management Simple and Standard. Supply Chain
Management: An International Journal, 2(4), pp. 124-
Mukhopadhyay, T., Kekre, S. & Kalathur, S. 1995.
Business Value of Information Technology: A Study
of Electronic Data Interchange. MIS Quarterly, 19(2),
pp. 137-156.
Pham Thi, T. T. & Helfert, M. 2007. Modelling
Information Manufacturing Systems. International
Journal of Information Quality, 1(1), pp. 5-21.
Ray, H., Taran, W. & Miyuki, S. 2003. Dating Example
for Information Architecture. CHI '03 extended
abstracts on Human factors in computing systems. Ft.
Lauderdale, Florida, USA.
Riva, C. & Rodriguez, J. V. 2002. Combining Static and
Dynamic Views for Architecture Reconstruction.
Software Maintenance and Reengineering
Proceedings. 6th European Conference on.
Rood, M. A. 1994. Enterprise Architecture: Definition,
Content, And Utility. Enabling Technologies:
Infrastructure for Collaborative Enterprises
Proceedin., 3rd Workshop on.
Rosenfeld, L. & Morville, P. 1998. Information
Architecture for the World Wide Web, O'Reilly
Associates, Inc.
Sherman, C. 2002. The Invisible Web, NJ, Information
Today, Inc.
Wang. R. Y. 1998. A product perspective on total data
quality management. Communications of the ACM,
41(2), pp. 58-65.
Wang, R. Y., Storey, V. C. & Firth, C. P. 1995. A
Framework for Analysis of Data Quality Research.
Knowledge and Data Engineering, IEEE Transactions
on, 7(4), pp. 623-640.
Wurman, R. S. 1990. Information Anxiety. New York.
Zachman, J. A. 1987. A Framework for Information
Systems Architecture. IBM Systems Journal, 26(3),
pp. 276-292.