|
Designing an Integrated Enterprise Model to support Partnerships in the Geo-Information Industry
New research perspective
We therefore focus our research towards the unbundling of the functionalities of current stand-alone systems in the traditional Geo-organisations, including mapping agencies, to make them available as independently developed, yet interoperable autonomous services. These functionalities include processes from different data sources, processes to create databases and manage their access, processes for map visualisation, GIS functionality for spatial data analysis, etc. Adding to these functionalities, the commercially available Internet-based services (Web Map Services), ready to use via standardised access interfaces and delivered across a wide area network; the OGC Application Service provider is an example. An infrastructure, with institutional and technical arrangement will be required to support the networking and chaining of these functionalities and services to create customised solutions and achieve common business goals. Integration is not limited to data exchange capabilities, but also concern with the rest of the enterprise by connecting all necessary functions and heterogeneous functional entities: business processes, information systems, application packages, organisation units, resources, etc. The infrastructure will manage information-, processes-, control- and work ows across the boundaries of the participating organisations. Basic functionalities and services are processes that run on Web services, made accessible via standardised, interoperable, access interfaces, semantic unification must be assured to support data exchange and need to be registered in order to be located in a distributed environment; all these are the main requirements to utilise such an infrastructure. Further, special services will be developed to provide the option of combining and chaining of services (a kind of broker/mediator), also to manage inter-organisational work ows and manage the quality of services in such wider network of services, operating under different rules and constraints. 4 From GDI to GSI Geo-information Data Infrastructure (GDI) provides access to geographical data by networking geo-information databases ruled by sharing mechanisms, defining technological as well as organisational aspects for the exchange of data. The role of the GDI is currently changing, from it being a simple data discovery and retrieval facility to become an integrated system suitable for the provision of customised information and services. For the sake of simplicity we use the term services to denote geo-information services. Normally developers address the issue of designing complex services by stringing together groups of functions in an ad-hoc manner. This approach may satisfy a particular need but doing this separately for different services hampers reusability. Moreover, lack, of descriptions of the solutions obtained makes it hard to aggregate solutions to execute complex tasks. Research is therefore focusing on the development of mechanisms to describe, combine and manage indepen-dent collections of services. Here we introduce a concept that aims at facilitating the generation of sophisticated value-added services. We call it the Geo-information Service Infrastructure or GSI for short (see Figure 1). The idea of the GSI is that elementary services can be described, accessed, combined and managed to deliver complex content. Within the GSI a common method is used to describe elementary services and their interfaces, and then these services are made available for users to create service chains that perform complex geo-processing tasks [12, 6, 13].  Figure 1: The GSI system concept A GSI is a system from which specialised geo-information products and services can be obtained by exploiting the artefacts of an infrastructure of interconnected nodes that include, among others, data repositories, data brokers, service providers, service brokers and clients. Large geo-processing tasks are achieved by combining or chaining artefacts located along the distributed nodes. Such combinations of artefacts provide diverse functionality that satisfies particular sets of require-ments. Every artefact has an economic value; these artefacts are assembled to perform operations within the infrastructure, resulting in a specialised artefact that has a value equal or larger than the value of the artefacts used. This architectural approach can be regarded as a “value-added system”. By chaining these artefacts one can provide a service. A service is defined as a behaviour of value to the user, which is accessible or instantiated through interaction points. This behaviour is exhibited through an appropriate combination of elementary artefacts. In order to bind multiple artefacts into a chain that accomplishes a large geo-processing task, a proper description of the participating artefacts is required. These descriptions focus on exposing the artefact’s internal behaviour, its intended effect and its interaction points or points of composition. These descriptions, which are presented as instances of well-defined models, make it possible to interchange and reuse artefacts. We call these descriptions system metadata and they are stored and accessible through a service repository. The GSI system enables Geo-Service Providers (GSPs) to make use of each others functionality to supply a wide range of services and possibly to reach larger groups of users. Figure 1 illustrates the interactions that take place as GSP nodes provide services to their users. Users interact with the different GSP-nodes to request their specific services. Figure 1 shows these interactions as dashed-lines. GSP-nodes may make use of artefacts available in other GSP-nodes in order to realise a particular service. These interactions between GSP-nodes are shown in Figure 1 as solid lines running from Node to Node. All connections mentioned here are established through a network. At the bottom of Figure 1 we can see that additional data collections located at non-GSP nodes may still be accessed, if needed, either by users or service providers. This is achieved by making use of the conventional data discovery functionality, of the clearinghouse server. These interactions appear in the Figure 1 as dasheddot-lines. |