Enterprise integration in e-government
The Authors
Raymond Cheng-Yi Wu, IBM Software Group, Singapore
Abstract
Purpose – The main purpose of this paper is to research the e-government architecture issue as a means of addressing the interoperability gap. It will incorporate the new knowledge acquired in the e-government integration methodology of the interconnectivity of cross-layers. These layers include the vertical direction of strategy, business, process, service and information.
Design/methodology/approach – The methodology used in this research as it progressed was component-base driven and involved the exploration of vertical e-government integration and common service interoperability between business and IT. The realization of this methodology first required a technical foundation setup followed by a business semantics study and a new concept of enterprise integration. The lack of interconnectivity between these layers is the main concerns with the implication that e-government architecture needs a robust micro-mechanism of semantic messaging and metadata to coordinate across layers.
Findings – E-government architecture became pervasive in the twenty-first century due to its significant growth in terms of huge volume transactions, the citizens' new service concept, and sophisticated businesses. Enterprise architecture mainly mediates between business and IT to minimize the gap by improving governance, agility and business integrity. All of these disciplines and principles should be applied to attain e-government transformation and vertical interoperability and common services provisioning are the major findings in the research.
Originality/value – This paper contribute new concept of enterprise vertical integration in e-government. The integrated solution of coherence of approaches forms the basis of this new concept; it serves as the backbone in vertical integration and addresses the e-government enterprise gap. It is expected to provide further insights into micro process integration across e-government enterprise layers.
Article Type:
Research paper
Keyword(s):
Government; Communication technologies.
Journal:
Transforming Government: People, Process and Policy
Volume:
1
Number:
1
Year:
2007
pp:
89-99
Copyright ©
Emerald Group Publishing Limited
ISSN:
1750-6166
Introduction
We all understand that in this new computing age of large-scale, high complexity and dynamic development, a common weakness in the interoperability between business and IT is leaving IT far behind in light of the direction business is taking. Poor business responsiveness and current IT governance make it even more difficult for e-government to achieve the enterprise goal. We identify this overall tactical issue emitting from the vertical gap as the “invisible enterprise issue” within e-government architecture. Finding new methods to minimize the gap between business and IT by using vertical enterprise integration is the purpose of this paper.
In this century, e-government transformation is one of the biggest challenges within the IT-related sector from the perspective of scale and complexity. The main objective is to adapt existing e-government to new computing requirements based on the citizens' new service concept. Thus, increasing service level features and quality as well as maintaining a high consistency of policies and rules, mission critical and time to services, interoperability and information sharing, productivity and efficiency in addition to reducing operating expenses are all priorities in this transformation. This will require multi-dimensional aspects from business innovation, service strategy, solution architecture, implementation and integration.
Figure 1 shows the evolution of the challenge; our goal is to lessen and counteract this challenge. In the past five years, SOA (Service Oriented), MDA (Model Driven), EDA (Event Driven) and many other leading technologies have been launched for the purpose of solving related problems. Their solutions have already given us significant improvements in e-government; however, the gap is still too massive to cope with the challenges of large-scale and high complexity over the next ten years.
Such a transformation is crucial as the service concept is changing rapidly; the coherence of the roadmap in e-government transformation is critical as most of the key service providers and vendors still lack a meta-enterprise vision in moving silo into a global roadmap.
The assumption from the discussion to this point is that the e-government issue is not merely caused by enterprise architecture misalignment between business and IT, but that a special style of architecture is needed to cover non-technical areas in order to streamline business implementation.
Methodology
This integration research was focused on the existing e-government issue which is the weak interoperability and misalignment between business and IT. The research included an investigation of the root issues, the impact analysis, the nature of interference between business and IT, a conceptual resolution and ended with a strategic direction. The methodologies used in this paper included component-base driven, vertical integration and common services. The proposed roadmap for e-government transformation was planned as follows.
First, in the pre-build stage – decompose the business block into business component; move forward to a service component through the business modeling process; the reason componentization needed to be set first because visibility and traceability are pre-requirements before vertical integration and common services, i.e. the new concept of interoperability needed to be a robust communication through integration and services; the concept of communication within e-government interoperability needed a service layer across a meta-level component rather than system to system or data to data integration. In this step, we, therefore, used componentization to build up the foundation for integration and services.
Second, in the building stage – after the realization of componentization, the other new concept – vertical integration was launched to work on the interoperability and fragmentation issues which are the crucial issues in e-government. The integrated solution of coherence of approaches forms the basis of this new concept that acts as a backbone in a vertical direction and addresses the e-government enterprise gap; it is expected to provide further insights into micro process integration across e-government enterprise layers. The three different types of repositories (metadata, services and semantics) collaborate with each other; the strategy of “enterprise vertical integration” is based on the processes that build up these repositories in the early stage as well as the methods for achieving interoperability between repositories. The symbolic meta-level repository integration is shown in Figure 2 and clearly demonstrates how the repositories can be integrated within an enterprise architecture environment.
Third, in the post-build and service provisioning stage – common services and common information are the values gained from the vertical integration across organizations; thus common services can build a framework from a virtual infrastructure for an organization to “plug and play” if the organization wishes to save IT cost and follow 100 percent business compliance. An organization also has the option to start from an earlier stage and use the service specification to develop its component within this common framework. As to the common information, this mainly consists of data model, meta-level specification, and the data access and information format. The disciplines are not limited only to the data itself, the interoperability between service and data are the key to provisioning common service and information infrastructure.
Discussion
Among the e-government challenges, interoperability has been identified as the most crucial. This is because the high integrity of non-system related policy and rules need to be followed across many geographical entities. Thus, a high semantic requirement, robust service interfaces and coherence of enterprise strategy are all important. Moreover, time to services for those mission critical, single source of truth of citizen based data as well as e-government interconnectivity to other service industries are all critical; consequently these requirements reinforce the need for a fundamental change of e-government. These called-for changes can be summarized as follows:
- Componentization, as mentioned above, became a trend in logical design to cope with the citizens' new service concept and mission critical requirement; these either newly-added requirements or existing legacy are moving forward to a component base architecture. There is a need to proceed through the processes of identification, specification and validation (Butler et al., 2003).
- Global-local architecture is the other trend in building up a topology foundation to achieve a common context, although the local authority may add their special rule leverage to global common practices. Within this topology, the gap between organizations is eliminated, as once service is componentized and the concept of government entity is remapped to this new topology, the government organization is virtualized. At this juncture development becomes faster as common services provide a pre-build technology foundation. Indeed, the fact that all of the organizations share the same service means they are all managed under the same rules which were developed using the top-down approach. As all of the organizations share a common protocol to communicate their semantic rules, the integration effort can be eliminated and interoperability can be enhanced through this topology once the semantic communication has been fully developed.
- Meta-level object definition and single source of truth are priorities in building up an efficient e-government architecture as the integration trend in IT has been changed from peer-to-peer (n:n) to hub and spoke (1:n), and from hub and spoke to service bus (1:1). The pre-requirements for this are the consolidation of all duplicate object and data fragmentation by means of aggregation, merge or summary toward achieving a meta-level. Thus, interoperability can be attained within the meta-layer to impact the individual entity effectively.
When we take the top-down approach into consideration, the strategic layer (i.e. an EU organization) has the authority to validate the rule within a physical domain and implement the business service through domain analysis; the process is completed by the decomposition of the context into a semantic candidate list. Current UK Government guidelines provide a metadata standard which can be a good reference in producing a candidate list for further micro process integration and interconnectivity between components.
Micro process integration was initiated by the domain context analysis process. In their domain analysis, Lee and Lee (2006) state that once sets of use case scenarios are formulated, the next step will be the extraction of verbs and nouns for identifying domain objects. The domain analysis concept was applied in supporting the theory of vertical integration in semantic analysis. When combined with the extraction of nouns and verbs, these make up the common components in process, object and interface as domain assets (Lee and Lee, 2006).
The pre-build semantic repository utilizes business domain analysis output in further development. Typical semantic development includes a list of candidate semantic objects. Thus, in addition to the semantic content, meta-level choreography of the semantic interface between different physical domains is also a concern. The semantic interface stored in the service repository needs robust communication with the information metadata repository in order to streamline cross layer implementation. Therefore, interoperability capability in a vertical direction allows mediation computing entities such as messaging carrier and class mapping (i.e. XML and JDBC) choreography between service and information to become easier and leading toward a seamless vertical integration (Valkeapää and Hyvönen, 2006).
Taking as an example its application in the citizen smart card scenario we see that “create a citizen smart card account” is a business service which consists of four composite services – “citizen medical history,” “security clearance,” “banking records” and “common application interface.” If we take “banking records” as an example for further decomposition to atomic services we see that these include “loan request,” “channel processing,” “credit and balance check,” “floor limit and fraud check,” “provision service and billing profile,” and “credit grant and adjust floor limit.” All of these services can be analyzed for their semantic objects. The use case analysis generates a descriptive sentence of these composite services which reads as follows: “a citizen submits a loan request for housing” (through form or web); “administrator processes application”; “authority processes check and credit review”; and “authority grants and adjusts credit.” The semantic analysis extracts nouns and verbs from atomic services and forms lists such as this. The semantic repository noun list stores objects such as citizen name, id, phone, account number and an attribute object such as status, credit, authority, and balance while the verb list stores check (balance, credit, floor limit); provision (service profile, billing profile); grant (credit [level 1,2,3]; authority [level 1,2,3]), etc. The metadata repository requires a pre-build data schema, data model, symbolic picture of data aggregation and business intelligence to transform the high human-involved development effort to simple configuration. As stated above, the realization of this concept is through interoperability between the service and metadata repository (Tannenbaum, 2001).
As mentioned above, common service and common information constitute the values proposed through the integration of semantic, service and metadata repositories under a global-local topology; this means that each individual organization no longer needs to create from scratch. Rather it can adapt its individual entity to the best means, within this common service and information conduit, to build the architecture from the portfolio level and communicate with each other by using a common service approach. The concept of virtual organization is to prevent any binding or constraint and to adapt the government organization to a shared common service and information based on demand. This feature enables a flexible topology to provide fast development options for “plug-and-play.”
Common service and common information is a top down approach which follows the nature of business implementation direction. As seen in Figure 3, from the technical perspective the idea was to eliminate the integration effort by using the “virtual organization” concept. From the organization perspective, this enhances the flexibility of service provisioning and the agility between organization and service; thus, the interoperability across organizations is increased. The realization of this top down vertical integration approach requires robust middleware to support the inter-communication between component-and-component and component-and-repository. In other words, the concept needs physical support from an integration service bus such as the middleware at run-time services.
M. Janssen and R. Wagenaar, in their “Towards a flexible ICT-architecture for multi-channel e-government service provisioning,” gave the following business scenario: a citizen goes to the counter of the town hall and requests a (new) driver's license (Figure 4); the process requires re-engineering and design guidelines which can be summarized as:
- information should be captured only once at the source and re-used by other components;
- communication between components should be kept to a minimum;
- there should be a central process-control component integrating business process steps with functionality provided by the components; and
- components should be able to communicate with each other directly or through middleware (Janssen and Wagenaar, 2003).
From the above design guidelines, the assumption can be made that citizen data integrity, componentization and centric semantic implementation, service interface and business process within municipality systems are all keys to e-government transformation. This implies that tactical e-government issues including data fragmentation, duplicate process and silo, multi-points integration and inefficient use cases are all within the scope of the top down approach. This approach consists of componentization, vertical integration, common services and semantic implementation (which is still new in e-government architecture and calls for further investigation). The proposition is that the following procedures need to be implemented before moving from silo to global mainstream: first – the semantic metadata format and protocol need be unified and synchronized in relevant geographical locations; second – terminology, natural language, semantic attribute and the access method need to be standardized in order to eliminate the semantic communication gap between two semantic groups.
In addition to semantic implementation, a service interface needs to be established to reach citizens. Thus, before service can be reached through channel and pervasive technology, the information foundation needs to first be consolidated for the purpose of building up a master data identified as “single source of truth.” This should be established through data identification, consolidation, cleansing and validation processes so that information can follow both a global-local structure and principles such as single point data manipulation and channel communication elimination, etc. Once a citizens' master data is built, service intelligence can be implemented to add value through service consumer insight and provisioning macro-trend analysis and personalization. Once in place, this foundation would provide an efficient interface to reach citizens through channel and pervasive technology.
To expand the service flexibility across organizations, a virtual organization concept was introduced as mentioned above. The e-government virtual organization presented in this paper is intended to highlight the service establishment through a combination of services provisioned by virtual organizations that leverage the policies, regulations, services and decisions into a single point within the vertical skeleton to facilitate cross organization services and overcome time and space limitations. In contrast to system integration, top-down vertical integration starts from context and is followed by service and information metadata to facilitate vertical integrated solutions and provide citizens with common service and common information. Figure 5 shows the integration of three repositories in global-local topology. The cross-organization integration known as integration service bus is seen as providing a physical connectivity service such as middleware.
Preston (1999) argues that virtual organization adds value and requires a different way of perceiving the world by those who wish to participate in it. Elements of virtual organization include the following:
- the relationship of a virtual organization is based on a particular competency that complements the others;
- it capitalizes on the mobility and responsiveness of telecommunications to overcome problems of distance;
- timing is a key aspect of relationships, with individuals using responsiveness and availability to decide between alternatives; and
- there must be trust between individuals separated in space for virtual organization to be effective (Preston, 1999).
Virtual organization can be distinguished from hierarchical and network forms of organization by its rejection of status boundaries and the lack of importance it ascribes to proximity. Formal hierarchies may connect distant offices and plants in a flow of information and material resources, but they enforce distinctions between departments as part of a system of top down command and control (Bush and Frohman, 1991) and rely on the proximity of management and workers within departments and offices to develop the trust necessary for a functional system (Handy, 1995).
This relationship and trust can be established under a top down approach by using global-local topology and vertical integration as opposed to peer-to-peer integration. The latter is limited in supporting relationships for the formation of a virtual organization which requires a particular competency within a group, IT loss mobility, and responsiveness and timing from the complexity and agility perspective. For this reason, peer-to-peer integration within e-government cannot be an asset from the virtual organization point of view. From the industry common methodology perspective, it is not difficult to discern that the concept of vertical integration was already in a prototypical stage although the trend had not yet been specified. The evidence for early signs of this new concept may be seen in the following:
- Zachman was almost the frontier of the enterprise framework; it never focused on any one architecture layer only, rather it introduced cross-mapping between the dimensional matrixes and was considered the first indication that vertical integration existed.
- Rational Unify Process (RUP) was used as a methodology of abstraction in identifying a boundary to look at use case in each small derivation and iteration and guided its architecture development in a vertical direction.
- SOA and MDA (model driven architecture) were architecture styles providing further indication that “loosely coupled” was actually a vertical alignment of business and IT as a foundation for the development of further business lines.
D'Souza (2001) defines vertical integration in his MDA (Model-Driven Architecture) and Integration Opportunities and Challenges and points out that vertical integration came from the different levels of abstraction within the same system, physical data models, logical data models, networks, application specifications, component assemblies, business process models, and business goals and strategies.
This viewpoint provides further evidence for the feasibility of the top down vertical integration. From the nature of the business implementation direction we have several options for the enhancement of interoperability. These include A2 A (application-to-application); B2B (business-to-business); D2D (data-to-data); or vertical B2A2D (business-to-application-to-data). If we start from A2 A or D2D, the risk is much higher than B2B because reverse engineering in business process re-engineering (BPR) requires both upward and downward re-engineering and n:n repeat integration efforts for both cases; this can cause a high risk if the change started from the business layer and each system was developed in a different scenario. B2B has lower risk concerns but is weaker due to its having a lower sharing model and component. By comparison B2A2D can have better business compliance across organizations with higher sharing common components, services and lower risk; however, the technologies cover many advanced areas such as semantic transmission, grid concept, componentization and service choreography, etc. which need further review. For example, if we assume we have most of the technology ready for this interoperability by using vertical integration and a top down approach, then even a minor weakness in the componentization of a business area that needs to apply a new policy change or organizational restructuring, with the potential for creating a big gap from “business change” to “common service” within this area, can cause delay or defect. The most likely reason for these setbacks would be that the method or componentization tool is not mature enough to achieve a full “separation of concern” level. Another area of concern would be the mapping between the semantic, service and metadata. Thus, bringing broad technologies and concept coherence to full realization needs further meticulous research.
Layne and Lee (2001) have the viewpoint that e-government provides progress toward a higher level of integration and interoperability to government levels and branches. In essence, interoperability leads to extensive information sharing among and between governmental entities. However, the obstacles preventing rapid progress in that direction are not merely technical. In fact, the technology side may prove the least difficult to address; the organizational, legal, political and social aspects may prove to be much more of a challenge (Layne and Lee, 2001).
This finding implies that the first steps toward interoperability cannot be seen merely from technical aspects such as A2 A, D2D or componentization of SOA. Rather it can be a high risk from a recursive re-engineering requirement through each business change caused by either government organization or political issue. Indeed, the political and organizational aspects are more challenging areas than technology. The maturity of the vertical integration concept and provisioning of cross-organization share service and information have a better chance to cope with both the political and organization issues and in eliminating technical integration complexity. This is evident from the nature of business direction, virtualization and interoperability perspectives.
Conclusion
The concept of vertical integration in e-government is achieved by applying semantic, service and metadata repositories to a global-local architecture foundation. The top down vertical integration methodology follows the nature of business implementation direction and has a better chance of coping with interoperability requirements. Basically, it can eliminate the point-to-point integration effort and risk as well as providing a common service and common information value proposition such as portfolio infrastructure for quick development and policy compliance.
The nature of “e-government vertical integration” distinguishes itself from traditional integration by effective transmission from business context and semantics into common components to produce “run-time” services and communicate with metadata through the micro-componentization process in a development environment.
The lack of interconnectivity between context, semantics, service and data are the main concerns with the implication that e-government architecture needs a robust micro-mechanism of semantic messaging and metadata to coordinate across layers. In this way, the strategic direction will have an opportunity to impact the business process to streamline successful services and information components.
Figure 1The IT evolution and requirements in new age
Figure 2Conceptual ontology interoperability between layers by using “semantic injection” and “metadata”
Figure 3Organization integration transformation through common services and information
Figure 5Differentiators of vertical integration from conventional integration
Figure 4Business scenario (citizen's application for driver license)
References
Bush, J.B. Jr, Frohman, A.L. (1991), "Communication in a ‘network’ organization", Organizational Dynamics, Vol. 20 No.2, pp.23-36.
Butler, J.C., Mayo, D.R., Weiler, J. (2003), "Succeeding with component-based architecture in e-government", FEA-PMO, .
D'Souza, D. (2001), Model-Driven Architecture and Integration Opportunities and Challenges, Kinetium, .
Handy, C. (1995), "Trust and the virtual organization", Harvard Business Review, Vol. 73 No.3, pp.40-50.
Janssen, M., Wagenaar, R. (2003), "Towards a flexible ICT-architecture for multi-channel e-government service provisioning", IEEE Computer Society, Proceedings of the 36th Hawaii International Conference on System Sciences (HICSS'03), .
Layne, K., Lee, J. (2001), "Developing fully functional egovernment: a four-stage model", Government Information Quarterly, Vol. 18 pp.122-36.
Lee, H., Lee, J. (2006), Analyzing Business Domain: A Methodology and Repository System, Korea Advanced Inst. of S&T, Seoul, .
Preston, S.M. (1999), "Virtual organization as process: integrating cognitive and social structure across time and space", Journal of Computer-Mediated Communication and Organization Science, .
Tannenbaum, A. (2001), Using Metamodels, Repositories, XML, and Enterprise Portals to Generate Information on Demand, Addison-Wesley Professional, Boston, MA, .
Valkeapää, O., Hyvönen, E. (2006), "Semantic annotation with browser-based annotation tool", paper presented at the first Asian Semantic Web Conference, .
Further Reading
Curtis, K., Foster, P.W., Stentiford, F. (1999), "Metadata – the key to content management services", Proceedings of the Meta-Data, the Third IEEE Meta-Data Conference, .
Maedche, A., Staab, S., Stojanovic, N., Studer, R., Sure, Y. (2001), Semantic Portal – The SEAL Approach, MIT Press, Cambridge, MA, to appear in: Creating the Semantic Web, .
Corresponding author
Raymond Cheng-Yi Wu can be contacted at: raymondw@sg.ibm.com (rwu_academic@yahoo.com)