一种分布式智能多主体决策支持系统模型

现代大型企业的经营活动通常分布在很大的地理范围内，甚至分布于整个世界。而在诸如商业、金融、管理等复杂领域建立决策支持系统(DSS)又需要综合考虑各种数据和知识模型，以及在该领域中发生的专业认知过程的推理模式。由此，本文提出一个分布式多主体DSS(MADSS)结构，它整合了不同类型的知识表示，并且具有任务导向的学习功能。在这个模型中，此系统由分布在各地的本地子系统组成，而子系统是基于多主体技术建立的，具有独立性和协作性。子系统由各种具有不同功能和能力的主体组成。主体具有基于用例的推理(CBR)能力和任务导向的学习能力，后者主要是通过KDD技术使其及时的在企业的数据库中挖掘有用的规则，并且通过知识评价机制提升学习结果的正确性。最后，我们用一个银行贷款受理的例子来说明这个模型。

A Model of Distributed Intelligent Multi-Agent Decision Support System

Modern large enterprises often spread their activity to many places,even around the world. And to build DSS in complicated areas will involve various data and knowledge types, and the reasoning mode in this area. So, this paper proposes a Multi-agent Decision Support System (MADSS) model that integrates different knowledge types and has a task-driven learning ability. In this model,the system is composed of distributed subsystems, which are based on multi-agent technology and are independent and cooperative. The subsystems are composed of various agents of different functions and abilities who have CBR ability and task-driven learning ability which is based on KDD technology with evaluation mechanism to enhance its learning.Finally,we illustrate the model with a bank loan case.