Study And Development On Integrated Intelligent Decision Support System Based On Data-driven And Its Generator

The new decision support technology such as data warehouse (DW) and data mining (DM) was born to offer new idea and methods for acquiring information and knowledge from data, and to extend the approach of in-depth information analysis for the manager. In the paper, the layer model of intelligent decision support system is studied and analyzed based on the study of traditional intelligent decision support system (IDSS), and the framework of integrated intelligent decision support system based on data-driven (IIDSS/DD) is put forward, and the data mining framework and its model are presented after analyzing of data character. What's more, theory, technology and methods on IIDSSG/DD are studied systemically and in detail and systemically. Finally, combined with actual enterprise, the special decision support system is developed to support the enterprise decision. Main works and innovations in the paper are as follows:1. The layer model of intelligent decision support system developtment is the theoretic basis of describing, developing and evaluating integrated intelligent decision support system based on data-driven (IIDSS/DD). Firstly, according to the process and character of the intelligent decision support system development, it is compartmentalized into such four layers as application layer, task layer, logical layer and physical layer, and each layer is described and analyzed. Secondly, in realization methods of task layer, the resource link layer is designed in detail and four system resources, such as data, model, knowledge and instance, are regulated in the two parts: BNF link language and the standardization of system resources link, with the use of some technologies.. Finally, with this understanding we study the framework of IIDSS/DD, which integrates such new technologies as data warehouse (DW) and data mining (DM) and componentstechnology and has the properties of generality, integration and extensibility, etc.2. The paper studies decision support technologies based on data-driven, which includes DW technology and DM technology, and discusses DW design methods and the data model of online analysis process. And then the paper puts forward and discusses the framework of data mining system development and the data mining concept model and analyses their relations. What's more, the actual functions of data mining are divided into several parts. Of them, the core of all the functions is the intelligent middleware which is made up of central process unit, assistant choice, database flat and functions combination flat. Finally, the paper studies in detail on material decision support technologies and methods, such as classification analysis, association rule mining, clustering analysis, etc.3. Combining traditional IDSS and new decision technologies, the paper analyzes and designes IIDSS/DD, including functions and work flow on data management, model management and knowledge management. The management of Data, model and knowledge is integrated organically to support in-depth information analysis and offer more valuable decision project in more broad fields.4. Based on object-oriented method and components technology, the component-based development method on intelligent decision support system generator (IDSSG) is discussed and used to design IIDSSG/DD, which is made up of such five components as main control component, model management component, data warehouse component, data mining component and knowledge management. Certain special IDSS can be developed quickly by using these components.5. According to the requirements of actual insurance, an intelligent client relation management system for insurance (CRMS/I) is developed quickly by using interrelated components of IIDSSG/DD. In the courseof development, components are the technical basis of system decision so that system functions are realized easily with the integration of data, model and knowledge. Based upon primary business system, the use of these components satisfis the actual requirement of enterprise insurance decision. The results show it is simple and valid that the component-based development method is used to develop the special IDSS.