Font Size: a A A

Research On Interface Extraction Mechanism And Fusion Method Of Multiple Decision Information In Knowledge Network

Posted on:2017-02-17Degree:MasterType:Thesis
Country:ChinaCandidate:Z Q DingFull Text:PDF
GTID:2209330485997824Subject:Management Science and Engineering
Abstract/Summary:PDF Full Text Request
In the current rapid development of information age, the decision-making is the core of management. Gradually diversified management disciplinary research makes the increased uncertainties involved in decision-making problem, and therefore to build an appropriate management mechanism to ensure the full diffusion and integration of knowledge is becoming an inevitable requirement of the decision-making smooth progress. And knowledge network is recognized to be the optimal solutions to these problems. The existing research results, however, did not directly on how to effectively extract and scientific decision information fusion method to research the precedent, only indirectly to mention the subject classification can provide decision-making information. People can actually get the decision-making information is often a lot of uncertainty and unstructured information form. It brings challenges to the decision makers to solve the problems of uncertainty and complexity.In this paper, based on the theory of interface management, the paper constructs the mechanism model of multi decision information interface extraction in knowledge network. And on the basis of the information expression mechanism, this paper proposes the method of information fusion based on ER rules, According to the ER rule, the uncertainty can be processed by the combination rule, which can be used for the fusion of multi-source information, and then get more comprehensive, more accurate decision-making information. In addition, due to the limited rationality of human minds and vagueness of the understanding question, we put forward the prospect of building methods. Main work and achievements are as follows:(1)In order to establish a knowledge network in decision-making interface extraction mechanism and model. The first step, through the basic concepts of knowledge network types, analysis of the basic characteristics of modern decision making, this paper proposed eight kinds of component models of the main body of knowledge. Then it describes the model from three aspects, which are physical realization layer, knowledge operation layer and environmental condition layer, which provide technical support and mechanism guarantee for the construction of the model. On this basis, we detailed analysis the four interface on the knowledge layer: diagnosis, creative, behavior interface, choose interface, and also analysis the information in each type of interface requirements, the body of knowledge type and technical implementation methods. Finally build body separation mechanism, decision-making mechanism of interaction and information expression mechanism, satisfied coordination mechanism to ensure the smooth implementation process of knowledge transformation.(2) Based on the information expression mechanism, this paper puts forward the method of information fusion. We combine the ER method of establishing rules put forward a prospect. First of all, according to the complex environment of incomplete experts concluded that susceptible to infer the characteristics of the information to build a multiple attribute decision-making model of information. Then based on the rules of ER to build steps of how to building the prospect, and compared this method with the prospect building method of single attribute decision making.(3) The case analysis. Using the mobile phone designing scheme selection of X group as an example. According to the above mentioned eight classes of knowledge subject, each choose one expert to evaluate all solutions and give the prospect of these schemes. After that, we use the method proposed in this paper to calculate the foreground information of each scheme. And then get the order of each program to provide a reference for the decision maker.
Keywords/Search Tags:Knowledge network, multiple decision information, interfaced extraction, Prospect theory, Evidential Reasoning
PDF Full Text Request
Related items