DESIGNING AND DELIVERING PUBLIC SERVICES
ON THE CLOUD
Yehia Taher
1
, Rafiqul Haque
2
, Dinh Khoa Nguyen
1
and Willem-Jan van den Heuvel
1
1
European Research Institute in Service Science (ERISS), Tilburg University, Tilburg, The Netherlands
2
Irish Software Engineering Institute (Lero), University of Limerick, Limerick, Ireland
Keywords: e-Government, Public services, Reusability, Customization, Cloud computing, T-Shaped platform.
Abstract: Cost and complexity are currently the most substantial obstacles for designing and delivering services in the
public sector. The traditional in-house development and maintenance landscape of public services require
experts from diverse domains, various technologies and complex on-premise infrastructure, etc. The high
upfront cost and complexity impede the proliferation of Information Technology (IT) within the domain of
public sector. It is the aim of this research project to deliver a cloud based platform that allows non IT-
experts to customize prefabricated and reusable public services by parameterizing them. This customization
revolves around reference guidelines that accommodate a methodology in a consistent manner.
1 INTRODUCTION
Recently, IT has undergone vigorous development to
facilitate organizations to develop and deliver
services in more efficient and effective fashion. In
this setting, public service organizations are
harnessing the power of IT and distributed
technologies (e.g., Internet) to automate their desk-
based human provided services and also to make
services ubiquitous. The ubiquity ensures that public
services are accessible by the users (e.g., citizen) via
the Internet. Despite enormous amelioration of
technologies, public service organizations have
failed to exploit the full potentials of IT. Our
analysis reveals three salient reasons behind this
including economical, technological, and societal. In
this article, we focus on economical and
technological. To be precise, this research deals with
two primary issues embodying cost and complexity
to design and deliver public services.
Traditionally, the in-house development
landscape includes a group of service development
experts from multiple domains (e.g., business,
technology), on-premise infrastructure, technical
specialists, and experts for maintenance of
applications and infrastructure. Evidently, such a
landscape augments cost and complexity
enormously that is not affordable and manageable by
a large number of public service organizations.
The above considerations bring the widely
known concept called reusability into the light of
this research. Reusability enables to reuse existing
elements that could fit into specific requirements. It
is a promising concept for developing public
services that do not require to be developed from
scratch. This helps to diminish the development cost
for public services. However, the only shortcoming
of reusable services is their fitness to specific
requirements because they are still generic (i.e.,
independent of any specific context). Hence, it is
highly unlikely that generic services can completely
satisfy the requirements of a specific context. This
implies that reusable services require
customizations. In order to customize context-
independent services to context specific ones,
organizations require experts from various specific
contexts (e.g. service usage context, service runtime
environment, etc.) and technologies (e.g., WS-*
standards, REST, etc.). In this setting, the reusability
concept turns out to be an effective means to reduce
the service development cost partially but not
completely. In addition, the licensing cost and the
cost of infrastructure increase the upfront cost
heavily that is beyond the ability of many public
service organizations to afford. Therefore, there is a
strong requirement for a solution approach that will
serve as a means of developing public services in
cost effective manner.
471
Taher Y., Haque R., Khoa Nguyen D. and van den Heuvel W..
DESIGNING AND DELIVERING PUBLIC SERVICES ON THE CLOUD .
DOI: 10.5220/0003393404710476
In Proceedings of the 1st International Conference on Cloud Computing and Services Science (CLOSER-2011), pages 471-476
ISBN: 978-989-8425-52-2
Copyright
c
 2011 SCITEPRESS (Science and Technology Publications, Lda.)

Figure 1: The traditional landscape of developing and delivering services.
Complexity is another grand challenge for public
service organizations to exploit the potentialities of
IT. Large numbers of organizations today are
striving to remove the excessive complexity from
the service development landscape. Many
organizations consume more time and effort on
maintaining their infrastructure instead of focusing
on adding value on services. Complexity may
underlie the entire infrastructure. The reason behind
this complexity is merely the integration of various
technologies. It turns the service development
landscape into a space only for experts who posses
sound knowledge on development platforms. In
addition, maintaining complex infrastructure is a
non-trivial task which requires a highly proficient
team of experts. These complexities preclude public
service organization to adopt IT for delivering
services to the end-users.
This paper aims at developing a cloud based
platform that will bolster the public service
organizations to develop and deliver public services
in efficient and (cost-) effective manner. We named
the platform T-Shaped platform. The T-Shaped is
grounded on the concept of reusability, a
methodology for customizing reusable pro cesses,
and most importantly the cloud computing
paradigm. This cloud based solution will facilitate
migrating the excessive complexity that arises on
account of in-house service development
infrastructure. This will encourage much wider
adoption of IT within public service domain,
promote the development of innovative public
services, and reduce the time to deliver services to
customer (e.g., citizen). The proposed platform
includes guideline that will enable the public service
organizations to develop and deliver services
without requiring experts. This implies, public
service organizations will be able to reduce the
service development cost significantly since they
will be able to lower their budget on experts.
Additionally, the pay-per-use economic model will
help reducing the cost of infrastructure and licensing
fees for technologies. In a word, the cloud based T-
Shaped platform will lessen the overhead cost of
designing and delivering public services.
This paper is organized as follows, a motivating
scenario is presented in section 2 to illustrate the
cost and complexity underlie in today’s solution
architecture; section 3 describes the core concept of
our proposed cloud-based platform for public
service organizations; section 4 describes the
methodology that facilitates the public service
customization using the T-Shaped platform. We
explain the related literatures in section 5 and the
future extensions of this research in section 6.
2 MOTIVATING SCENARIO
We present a motivating scenario in this section.
Figure 1 shows the scenario. The scenario
demonstrates various development costs that exist in
traditional development landscape. It also reflects
the complexities that are underlying the traditional
application development ecosystem.
Figure 1 demonstrates several areas of traditional
in-house service development ecosystem, which
CLOSER 2011 - International Conference on Cloud Computing and Services Science
472

increases the cost massively. The red-dotted
rectangles in figure 1 spot these areas. In the given
scenario, a public service provider develops services
that run on in-house platform. This requires (at least)
a team of developers lead by a specialist, analysts,
solution designer, and solution architect. Hiring
these experts increases the upfront development cost
exponentially. Besides, the platform is integration of
various technologies that also raise the cost
immensely. The reason is merely the licensing cost
of enterprise editions of these technologies.
Additionally, the public service providers host
the services on on-premise infrastructure that causes
a considerable escalation of the total cost. The
infrastructure entails various equipments in
particular, processing equipments (e.g., CPU,
memory), storage equipments (e.g., database
storage), and networking equipments (e.g., switch).
In addition, a group of technical specialists need to
assemble these equipments in an infrastructure. The
costs of these resources (equipments and human)
increases the total cost heavily that is far-affordable
to many public service organizations.
Furthermore, in-house service development
platform can be enormously complex. The platforms
today integrate several technologies that raise
complexity. The classic examples are Eclipse and
Netbeans platforms that are largely complex to
install and operate for a non-expert user. Developing
applications or services using these platforms are far
beyond their ability. Figure 1 shows that a traditional
development platform integrates different
technologies including Silverlight, Intalio BPMS,
and so on. These technologies make the platform
infrastructure highly complex that is not easy to use
for users who are not adequately expert. Developing
services using this platform by a non-expert user is
merely impossible. In addition, there are no clear
guidelines that can ease the service development
complexity.
Now, the above issues foster one important
question: what is the most suitable approach that
optimizes the total cost and reduces the complexity
of developing and maintaining public services? We
initiate this research to find answers of these
questions.
3 T-SHAPED PLATFORM AS A
SERVICE: CONCEPTUAL
OVERVIEW
Adopting the cloud paradigm, we offer the T-Shaped
platform as a service for public service organization
to design and deliver services in a cost-effective
manner. The T-Shaped solution aims at providing a
virtual development platform for the public service
administrators to streamline the public service
design, delivery and management processes on the
cloud.
In this section we describe the fundamental
concept of the T-Shaped platform. The T-Shaped
platform consists of two different views: Horizontal
View and Vertical View. For service design, the
horizontal view of the T-Shaped platform proposes
that public service administrators can reduce the
transparent cost by exploiting a number of generic
Reusable Services in the public service domain
together with a Reference Guideline for customizing
these services. The connector of T-Shaped platform
connects a provider to a public service repository
where providers can query and find reusable
services. The T-Shaped platform embeds a reference
guideline that underpins the customization of
reusable services.
Following the vertical view of the T-Shaped
platform, the users may use the reference guideline
to customize the generic public services as they
want. The reference guideline serves as the guiding
principle for public administrators or service
providers accommodating the customization of
generic services without the need of having intense
knowledge on processes as well as its related
technologies. It contains a large set of parameters
derived from diverse domains that are important to
the customization of public services. T-shaped
allows public administrators to use these parameters
to customize the process-based services. The key
idea of the reference guideline is to bolster stratified
customization which allows fine-tuning a context
independent service to context specific one that
represents the unique interests or characteristics of
an organization. Figure 2 depicts the stratified
customization approach with an example.
Stratified customization is an approach that
promotes customization of context independent
services to context specific ones in a multiplicity of
1 to n.
The meta-reference model in figure 2 is a
reusable process model that associates services
containing the elements from global perspective that
is independent of usage and context. The
customization produces the reference model
containing the elements and characteristics for
specific usage but not context and finally the
solution model that is the concrete solution to a
specific context. The example in figure 2
DESIGNING AND DELIVERING PUBLIC SERVICES ON THE CLOUD
473

Figure 2: The stratified customization approach with
example.
demonstrates the customization process. We provide
the example linked with customization approach
using dotted lines. The meta-reference model is the
permission process associating permission services.
The process is customized to permission process for
residence permit which is not yet a concrete
solution. The final customization produces the
concrete solution which is residence permit for
Tilburg Municipality. We discuss the methodology
for customizing services in section 4.
Developing services in-house may require much
work in addition to the actual service logic
development. This work may include the
procurement, installation, configuration, and
operation of platforms, servers, storage, and
networks as well as the provision of database
services and user-authentication mechanisms.
Providing these portions of the development process
as cloud platform services is expected not only to
reduce the development costs but also to ease the
work load of the public administrations while
improving the service value they provide. Inspired
by this idea, our T-shaped solution will be delivered
to the public administrations as a virtual
development and hosting cloud platform service.
Conventional on-premise service development in
the public service domain, as explained in section 1,
is unacceptable due to high complexity and
development cost. Moving to our virtual platform
helps to delegate all the complexity of the
development lifecycle of public services to our
platform. In particular, this enables public
administrations to better concentrate their efforts on
understanding the citizen requirements and
preferences which - without any doubt – leads to
more adaptive public services to the citizen ever-
increasing dynamic needs. We believe that this
approach helps much for creating public services
that continue to have a significant positive impact on
the citizen’s life. Furthermore, public
administrations, when using our T-shaped solution
platform, may be charged based on the pay-per-use
or subscription schemes with a variety of predefined
pricing models and agreements. That gives them the
possibility for considering the most economic way
of developing their public services.
In the next section, we describe the methodology
of our cloud-based T-shaped platform for public
services.
4 METHODOLOGY
The T-Shaped underlies a methodology for
customizing reusable public services. We propose
this methodology to facilitate stratified
customization of processes. The methodology is
influenced by the research works from (Ma &
Leymann, 2008), (Heuvel & Jeusfeld, 2007). It
includes three phases that are described in the
followings:
(i) Discover
The public service provider finds the required
services from the service market or service cloud.
The T-Shaped solution connects a provider with the
service cloud.
To support finding and loading of services from
a public service cloud, the following primitives need
to be supported:
• Find: The provider finds the service from the
service cloud using this primitive.
• Load: This primitive is used to import a service
in T-Shaped platform to customize to be used in
specific context.
Figure 3 shows that a public service provider
find and load public services from the service cloud.
Figure 3: Discovering public services from the cloud
service.
(ii) (Re)-Design
After the service is loaded on T-Shaped platform, a
service provider analyzes the loaded service. The
CLOSER 2011 - International Conference on Cloud Computing and Services Science
474

analysis fosters re-designing the service to satisfy
specific interests. Comparing with specific
requirements, the provider selects the list of
activities that should be refined, pruned, and
aggregated. The service may need extra
functionality to be extended to satisfy the context
specificity. To support public service providers, we
integrate the following primitives with the T-Shaped
platform:
• Refine: Refine allows refining an activity into sub-
activities. For instance, prepare permission
application is an activity may be refined to
prepare document and fill application.
• Prune: An activity can be removed from the
process using this primitive.
• Aggregate: Aggregation allows combining two or
more activities into single activity.
• Extend: This primitive used to extend the required
functionality.
Additionally, activities can be renamed to fit into
a specific context. Renaming is labelling the
activities, actors, events of the generic process.
Figure 4 shows how T-Shaped solution facilitates
a service provider redesigning a generic process
using the operators.
Figure 4: Redesigning a Generic Process using T-Shaped
Solution Operators.
(iii) Parameterize
After the service is scoped to a specific interest
through redesigning the underlying process, this
phase supports parameterization of the service.
Parameterization allows specifying the quality of
services (QoS), policies, and security parameters. It
also allows specifying the values of parameters.
Thus, parameterization is highly critical to reach the
goal of service and also to ensure the user
satisfaction.
In this regard, T-Shaped provides an
idealsolution, a reference guideline that delivers a
large set of parameters from diverse domains (e.g.,
regulatory policy) to support service provider to
parameterize services. The providers will be able to
select and specify the parameters without having
much expertise. Figure 5 shows an example of
parameterization of a service.
Figure 5: Parameterization of Public Services.
5 RELATED WORK
This research revolves around three key concepts
cloud computing, reusability, and customization. In
recent days, cloud based solution for developing
applications is gaining enormous popularity with a
high number of market-available platforms.
(Heroku, 2010) was one of the first cloud platform
providers, which delivers the whole lifecycle
management solution for Ruby & Rails application
over the Internet. Another prominent example in
cloud-based development platform is the Force.com
(Salesforce.com, 2010) that allows designers to
build their CRM applications faster and with lower
cost using a very simplified programming model.
Instead of providing a whole virtual development
platform like Force.com, (Google App Engine,
2010) and (Microsoft-Windows Azure, 2010)
provide SDKs that support a simulated development
and testing environment.
The existing cloud-based solutions available in
the market may largely contribute to reducing the
upfront cost of acquiring in-house development
platforms and the complexities that underlie the
traditional service development ecosystem.
However, these solutions are only provided as
generic, domain-independent solutions as they
mainly target a large number of clients. Yet in the
public service domain, expert cost regarding
developing services is still a question since public
service organizations may require a high number of
experts that have adequate domain knowledge. Our
DESIGNING AND DELIVERING PUBLIC SERVICES ON THE CLOUD
475

approach is certainly exceptional in this regard since
it particularly focuses on the public service domain.
Furthermore, the concept of reusability of public
service processes is heavily documented throughout
various bodies of literature. (Ma & Leymann, 2008)
proposed a research that highlights fragmenting a
complex business process into shards that are
intended to be flexible and reusable for future
business process modelling. This research work is
enormously interesting especially the life-cycle
model for business process modelling using reusable
fragments. However, the scope of the work does not
solve our problem entirely since we focus not only
on facilitating reusability but also a robust guideline
for process customization, which add value on the
top of reusable fragments.
(Curran et al., 1997) initially proposed reusable
business processes as an approach for large-scale
enterprises. Their work has been cited in an
extensive number of research works, yet was
criticized by (Mendling et al., 2006) with
counterarguments on SAP reference model. (Heuvel
& Jeusfeld, 2007) proposed a framework with
guidelines to transform a model with reference
models in particular, the SAP reference model.
Strictly speaking, these researches are only
conceptually related to our approach but different in
terms of applications. In this paper, we narrow down
our scope to public service reusability that has not
been considered yet. The closest work related to
reusable public service has been proposed by
(Koussouris et al., 2008), in which the authors
presented a modelling view of generic processes.
Their main contribution was to support the Public
Administrations to achieve resolution of
organizational interoperability and systematically
address the Homogeneous Service Composition.
Fairly speaking, there is no solution approach
available that entails the cross-domain parameters to
support multi-modal customization.
6 CONCLUSIONS
The lack of expertise, capital expenditures, and high
cost of experts are the predominant factors that
severely preclude wider adoption of eGovernance
among public service organizations.
In the paper, we have presented a T-Shaped
cloud based solution which aims primarily at
reducing the costs and enhance the efficiency and
effectiveness related to public service design and
delivery process. The T-Shaped is expected to
influence the public administrations to embrace
electronic governance. The proposed solution is
rendered as a cloud based platform focused to
alleviate the involvement of public administrations
with many aspects ranging from consultation,
design, implementation, operation, to maintenance
of applications and IT infrastructures. This enables
public administrations to better concentrate their
efforts on improving the value of the rendered
services by relying on the T-Shaped customization
facilities.
Our proposal reflects first attempt in defining a
foundation for designing and delivering public
services on the cloud. As part of our ongoing and
future work, we will conduct empirical/experimental
study focusing on the security and privacy including
data confidentiality aspects of public services
delivered on the cloud.
ACKNOWLEDGEMENTS
The work leading to this result is sponsored by the
EU FP7 as part of the COCKPIT project.
REFERENCES
Cabinet Office, (2005). eGovernment Unit: eGovernment
Interoperability Framework. Retrieved from: http://
interim.cabinetoffice.gov.uk/govtalk.aspx
Curran, T., Keller, G., and Ladd, A., (1997). SAP R/3
Business Blueprint: Understanding the Business
Process Reference Model. Enterprise Resource
Planning Series. Prentice Hall PTR,
Google App Engine (2010). Retrieved from http://
code.google.com/appengine/
Heuvel, W-J. V. D, Jeusfeld, M., (2007): Model
Transformations with Reference Models. IESA 2007:
63-75.
Heroku (2010). Retrieved from http://heroku.com/
Koussouris, S., Tsitsanis, A., Gionis, G. and Psarras, J,
(2008). Designing Generic Municipal Services Process
Models towards eGovernment Interoperability
Infrastructures, eJETA 2008.
Ma, Z., and Leymann, F., (2008). A Lifecycle Model for
Using Process Fragment in Business Process
Modelling. Proc. of the BPMDS’08.
Mendling, J. et al., (2006).
Mendling, J., Moser, M., Neumann, G., Verbeek, H.M.W.,
Dongen, B.F.V., and Aalst, W.M.P.V.D. (2006):
Faulty EPCs in the SAP Reference Model, in Proc. of
the BPM 2006, LNCS 4102, pp. 451–457, 2006.
Springer-Verlag.
Microsft-WindowsAzure (2010). Retrieved from http://
www.microsoft.com/windowsazure/.
SalesForce.com (2009). Retrieved from http://
www.salesforce.com/platform/
CLOSER 2011 - International Conference on Cloud Computing and Services Science
476