| Enterprise level Decision Support System(EDSS) is an intelligent system to support managerial decisions under the information environment within a whole enterprise, which employs the internal and external data sources as well as mathematical models, expert knowledge and other intelligent decision methods. EDSS has all the features of an Intelligent Decision Support System(IDSS), and has higher requisitions in results, utility of decision resources, and implement techniques. EDSS is enterprise oriented, aimed at multi-objects, and the integrity of qualitative and quantitative objects. The decision resources rely not only on expert knowledge acquired through knowledge engineering, but are relying more on the historic transaction data and extraction of knowledge automatically through machine learning and knowledge discovery techniques. Besides, EDSS is related to information system in every level, and should rearrange its decision style according to business changes, the diversity in decision methods and agility in its architecture is highly emphasized. EDSS should reflect the decision environments, objects and demands of enterprise decisions, fully employ the achievement of EDSS, and explore it's own decision processes, intelligent decision methods and architecture. These features are exactly what emphasized in this dissertation.This dissertation researches on the theory and methods of EDSS according to information architecture, information environments, information demands and information characters of managerial decisions. With emphases on the speciality of environment and the complexity of enterprise decision problems, it covers these aspects as information features, data driven decision making processes, fusion of decision methods and software architecture. It thoroughly explores into issues of knowledge acquisition, decision strategies, the compound of qualitative and quantitative decision methods, re-configurable architecture with agility and adaptability, etc. thus forms a preliminary principium framework of EDSS.Contributions of this dissertation include:1. Summarizing states of the art of EDSS, analyzing intelligent decision methods, and using the intelligent features as a main clue to review architectures, system models and future trends of IDSS. Challenges that faced EDSS researches are also discussed.2. Discussing features of enterprise level decision making. The reticulate relations among data, information and knowledge, information architecture and information environment, information acquisition processes, selection of strategies in information decisions are explored; Data driven decision making methods are studied, and A DataWarehouse based EDSS model is presented. It then probes into the knowledge process, representation and also the development of knowledge-based systems.3. After reviewing the applications of maximum theory in managerial decisions, it presents terms of "Intelligent adaptability degree" and "Compound Intelligent adaptability degree" to evaluate the suitability in selecting intelligent methods to certain problem and decision making environments, presents it's calculating method using fuzzy pattern recognition, and then discusses integration methods of optimal decision methods with artificial neural network and expert systems.4. Researching on the support of decision process in an EDSS. After summarizing the integration style of expert system and DSS, this dissertation presents a Multi-agent based architecture that supports decision making processes, specifies some kinds of Agent, discusses the collaboration and cooperation of agents, and describes a Multi-agent based capacity planning decision support system to illustrate the decision processes.5. It probes into the hierarchical architecture of information and decision support systems, and gives a formal specification of the components and their connection, discusses component based hierarchical architectural model, studies component based EDSS platform and it's learning and control mech...
|
PDF Full Text Request