نمایندگی فعال سیستم های پشتیبانی تصمیم گیری خدمت محور

The design of Decision Support Systems have recently emphasized web enablement as the next step in design improvements for this class of applications. We argue that these approaches fail to address the key notion of adaptability in the support for decision makers. Instead of focusing exclusively on automation in decision making, we believe it is also necessary to pay attention to the interplay between decision makers and organizational processes. The service oriented view of organizations recognizes the need to accommodate the changing reality of organizational dynamics. For example, the service science approach focuses on interactions between service providers, their clients, and consumers as important interacting components of a service system. Current approaches to DSS design are constrained in terms of their ability to adapt to changes in user requirements and to provide support for the evolution of systems. This situation worsens when resources are distributed at different locations across organizations, decision making processes are required to be integrated at different points in time, and when collaboration is needed among decision makers. However, this typically characterizes the needs of collaborative decision making in networked organizations as exemplified by systems used for supply chain management. To address these problems we leverage the power of services for designing a framework that explicitly recognizes the need for design based on service delivery. We develop an agent-enabled service-oriented architecture to realize the proposed framework with service and agent paradigms. The architecture is refined and validated with an implementation in the supply chain context.

Organizations, large and small, local and global increasingly define their primary purpose as ‘service’ rather than products [10] and [17]. In both global and local contexts, for firms of all sizes, the delivery of service increasingly relies upon both intra- and inter-organizational connectivity that includes social and technical connections, which are effective and efficient. Systems to support organizational decision making has been discussed in the literature for several years. Two important trends motivate us to rethink fundamental design issues pertaining to implementing such systems. First is the development of tools and methodologies that take a service oriented view of design. Business processes need to be supported by clearly acknowledging the need for a collaborative environment amongst organizational units to deliver solutions. A useful way to address this issue is to build the notion of service delivery directly into design implementations. Second is the recognition at an organizational level that service interactions are key functions of successful modern organizations — the ubiquity of supply chains in organizations being a good case in point. We discuss these issues and link them to the objectives of our work. Current research perceives service as the foundation for all economic exchange [40] and [41]. Markets and corporations have shifted toward a service-oriented environment, which raises the need for new approaches towards service research. Palmisano [33], Maglio and Spohrer [29], Spohrer et al. [37] and others have called for a “science of service”, aiming to provide new theories and practical advice for service firms. The role of information technology in this process has subsequently become a research priority for service science [32]. Every service system relies on the interaction and sharing of information with customers during service engagements. Customer involvement is generally seen as a prerequisite for successful service engagements and value co-creation. Driven by technological changes, service engagements have shifted into virtual realms, resulting in so called technology-enabled value co-creation. Traditional approaches to studying the services theme have differentiated tangible products from intangible services in order to clearly articulate the point of differentiation for items with economic value and their interaction with markets [27]. The emphasis of this work is on examining a particular characterization of service delivery — one that leverages the use of IT to achieve this objective. The particular issues that we attempt to look at revolve around a better understanding of business process definitions, distributed project compositions, and coordination amongst project teams. The phenomenon of service systems as enablers of business innovation has been discussed by various researchers. Conceptually, we can envision three components of the service framework. The first is a service provider that could be a combination of individuals, organizations, and technologies. The second is a service client that could be a combination of individuals and organizations. The third is the object or target of the service — the entity that will be transformed by the application of the service. This could be a combination of people, organizations, processes, technologies, information, etc. Fig. 1 below (adapted from the work of [29]) highlights the key aspects of thinking about service systems and the issues that arise when we want to study such entities.Such a framework enables us to define a set of interesting research issues around the interaction between the components of the framework, including an emphasis on intra- and inter-organizational connectivity. Organizational decision making is a dynamic and complex process. Decision making requirements are constantly changing over time and vary from person to person. They are made at different levels of an organization, and involve many people from different locations with different business orientations towards the decision making process [9]. This variation in terms of organizational level and orientation makes collaboration among the decision makers vital. The decision making process, whether it be structured, semi structured or unstructured, requires data and information from a multiplicity of sources. Decision makers need to quickly get to the essence of the underlying data/information by going through a gathering, accessing, integrating, transforming, discovering and learning cycle. The application of this cycle gradually fine tunes the decision scope either by continuous inputs over a period of time or through the decision maker's insightful inputs at specific points in the decision making process. Systems to support decision makers in the decision making process have been formally recognized as a separate class of systems since the 1970s. Decision support systems (DSS) [6] and [36], have been formally defined as interactive computer based systems [1], [4] and [38] that support decision making processes for decision makers to solve semi-structured and/or unstructured problems [19]. In the current organizational context, such systems need to be adaptive in order to obtain the various resources needed in the decision making process [2] and to cope with changing decision making requirements for different people. Adaptive systems must have the capability to support users at different locations to access various resources across networks to enable them to participate in synchronous or asynchronous decision making processes. Decision oriented systems must also provide iterative methods such as what-if analysis to help users fine-tune their decision making. They must be able to adapt to changes in the external, internal, and system environments, as well as in the users themselves. Furthermore, these systems should enable the flexible manipulation of components and processes. For example, there should be no restrictions imposed by the system on users for the selection/integration of decision making components, nor should there be restrictions on acquiring the necessary resources from heterogeneous distributed systems. Equally, the architecture should allow the addition of new components into the system at run time. The problem that we address in this paper therefore can be captured as follows. We propose that current organizational realities necessitate an extension of conventional thinking about decision support systems along multiple dimensions. First, collaborative decision making among organizations requires effective exchange of information among them. Second, such exchange can be usefully facilitated through proper systems design that views information exchange as the provision of appropriate services. Third, systems must be adaptive to accommodate the needs of dynamic and heterogeneous organizational entities. These problems and the issues that arise from them motivate us to propose a high level service oriented decision support conceptual framework. In this paper, we outline the development of an agent-enabled framework and architecture to support the service oriented requirements of modern collaborative organizations. We adopt a design science paradigm [20] to develop a framework and implement a prototype that emphasizes service delivery as a key guiding principle for DSS development.

Decision support systems are designed to support decision makers for solving unstructured or semi-structured problems in business organizations. Decision makers are often geographically distributed and their decision making can affect each other when solving problems within a complex environment. Many systems are deficient in providing support for such complex decision making because they lack the capacity of integration with other enterprise systems. These and other problems can be mapped into the issues of resource, location, lifecycle and time in DSS design. To address these problems we propose the abstract distributed decision support framework. Our framework is predicated on a contemporary interpretation of organizational work where collaboration between organizations is vital for the delivery of high level services such as data analytics. This abstract framework is able to adapt to changes in its environment to support any decision maker using any system platform, at anytime and anywhere inside/outside of an organization. This motivated us to propose a distributed decision making process to support complex decision making in organizations keeping in mind the four dimensions. A distributed DSS framework and an agent-enabled distributed DSS architecture were designed to support this decision making process. The proposed service-oriented framework was designed with the main concepts of components, mapping, DSS, SDSS and scenario in a distributed environment. Components such as data, models, solvers, visualization and scenarios are collected and integrated into decision support systems to support users (system builders and decision makers). DSS components and DSS are distributed across intra, extra and inter layers to address the resource, location, lifecycle and time dimensions. The ASDSS architecture was introduced to realize the concepts in the frameworks. Data, model, solver and visualization concepts were realized by component objects in an OOP paradigm. Mapping was realized with OOP technology. The intra and extra, and inter layers in a distributed structure were realized by the Intranet, Extranet and Internet systems. The agents were used in the ASDSS architecture to realize the concepts of resource, location, lifecycle and time dimensions. To support resource and location dimensions, the construction of agent shells and DSS components gives the agents the capacity to carry components. To support the lifecycle dimension, the agent lifecycle facilitates the component lifecycle. To support the time dimension, the agents work as a team in the agent community, sharing the objects for parallel or sequential decision making processes. A prototype in the CPFR domain was built to implement, assess and refine the process, frameworks and architecture introduced in this paper. A key issue that usually arises with respect to design oriented work is the validation or evaluation of the design framework. In this paper, we propose that the design product is a proof-of-concept of our generalizable framework for DSS implementation. Our design approach is built fundamentally on four key dimensions: resource, location, lifecycle, and time. For the resource dimension, our design shows that it has addressed the key issues of independence, integration, sharing and extensibility, all of which are essential in the collaborative service oriented organizational structure. For the location dimension, we focus on mobility, reach, and communicability as the key necessary functionalities for DSS design. For the lifecycle dimension, the design emphasizes scenario lifecycles, model lifecycles, and automation as important considerations. Finally for the time dimension, the design provides support for synchronous and asynchronous functioning as well as sequential and parallel execution. Section 7 provides one approach to validation through example where the various functionalities are demonstrated. This may be seen as one of the limitations of the work. Further validation may be carried out in simulated organizational environments or through the use of action research by embedding the implementation in an organization [20]. The design approach that we propose, while robust in terms of functionality, needs implementation based fine tuning before such organizationally based validation exercises are possible. Situating complex technology based organizational interventions in an organization is an elaborate but essential exercise; however in this paper, we offer the first step necessary before such an exercise can be undertaken. Our approach to design for supporting inter-organizational decision making offers potentially practical solutions for managing complex collaborative business processes. We do not believe that a technical solution alone will be successful in managing such processes. However, proper designs can go a long way in supporting decision making processes that are contemporary and relevant. This work offers a significant potential to enhance the effectiveness of the use of technology in a decision making context in dynamic organizational environments. While the notion of supporting decision makers has existed for a number of years, two fundamental and significant shifts have necessitated a rethinking of design principles of such systems. The first is the evolution of the underlying technologies that incorporate better design practices, tools, and methodologies and the accompanying growth of technological awareness of decision makers as individuals. The second is the service orientation of design methodologies that recognizes the complexities of modern organizations the need for collaborative design implementations. This is a reflection of the broader notion of collaborative organizations that are well served by a service orientation. Our work addresses the issue of decision support for organizations by incorporating the service orientation directly into design processes.