For Chinese Type Of Decision-making Problem Of Intelligent Recognition Method Research

Problem Management System (PMS) is one of the key subsystems of problem-driven decision support system (DSS). Its primary function is to understand and solve decision problems intellective.However, with the enormous development of science technology and economic society, various uncertainties make decision problems more and more complex, increasing the difficulty for DSS to process problems.And therefore people’s expectations of DSS have been reduced.So, it becomes the keystone of problem management method researches that how to improve the intelligence and correctness to understand problems for PMS.It is a process to understand problems by utilizing relate knowledge. Its finall purpose is to get the models to solve the problems by identifying their types.In the process, it is the most critical to identify the types of problems accurately because the problem type determine the solving model type. For this reason, it becomes the content of this paper that how to make the PMS identifies the problem type then the solving method type of a given problem quickly and accurately.According to the understanding depth, the identifying process can devide into two stages. In the first stage, the main task for PMS is to identify the similar problems by using sentence similarity computing and then using the method of the most similar problem as the method of the target problem. In the second stage, the main task is to analyze the ontic structure of the target problem deeply and then get the problem type and the method type.Starting from the two stages, some researches have been conducted on the basis of current researches.The main contents are as follows.(1) A new approach for similar decision problem identification is proposed.By analyzing the common characteristics of Chinese sentences of decision problems, this paper utilizes sentence similarity computing to study a more accurate method for decision problem similarity computing.(2)The ontic structures of decision problems are analyzed. Starting from the human cognition process based on concepts, this paper summarizes the common process of learning and understanding problems for human.Then the ontic structures of problems are analyzed and three kind of knowledge concept networks are presented.(3)Some Models based on ontology for identifying problem solve methods are proposed.Taking the case of programming problems and forecast problems, this paper presents the ontic rules and the reasoning processes for identifying the problem type and the solving method type.(4)The applications of the models in a DSS are discussed.Taking an enterprise for an example, this paper applies the models to a DSS in order to validate them.