Bayesian Networks Application Research In Adaptability Teaching System

In recent years network teaching platform increasingly emerge, but the personalization teaching of many network teaching platforms can not been proven satisfactory and network teaching quality is not high. Network adaptability teaching becomes the inevitable tendency of modern remote teaching development, student model plays an important role as the composition part of adaptability teaching system for developing the teaching of adaptability. Therefore research how to design and use student model to realize the personalization of remote teaching with adaptability, has important meaning for the research of remote educational field.The adaptability organization of teaching resource is the crucial content that carries out adaptability network teaching, adaptability teaching system need to organize adaptability teaching resource according to student model, through research thinks that definition of teaching resource according to the needs of adaptability teaching is helpful for the adaptability organization of teaching resource, that is feasible channel so that network teaching realizes a kind of adaptability teaching; at the same time,adaptability teaching system need accurately evaluate cognitive state about the content that student have studied, system organize the teaching content of adaptability according to student's cognitive state, so student model should contain student's cognitive state about studying content. Therefore this paper joins teaching resource and student model, student model is finally divided into the model related to field knowledge (teaching resource) and the model unconcerned with field knowledge. Since there is plenty of uncertainty factor and uncertainty information during student learning course, for building model this paper is considered to seek a kind of method and the model that can handle uncertainty knowledge.Bayesian network is now the most effective one of theory model in uncertainty knowledge expression and the inference field, and applies to uncertainty, probability knowledge expression and inference, it is that a kind of digraph based on network structure, can carry out two-way parallel inference, and can synthesize empirical prior information and posterior data, make inference result more accurate and credible. Therefore Bayesian network's application in adaptability teaching system has important meaning.This paper with pedagogy theory have first introduced Bayesian network theoretical foundation, then utilize Bayesian network to build student model: have described the course of model's detailed design, include the establishing of domain knowledge graph, the learning algorithm of condition probability table, evidence information getting as well as evidence spreading algorithm in network, the knowledge expression method of Bayesian network in student model have also been studied. During student's cognitive diagnosis process, with vague algorithm realizethe diagnosis of ability and evaluation, and evaluating result make Bayesian network get evidence information. Finally, an experiment system based on above-mentioned thought has been designed to verify teaching resource adaptability organization function. Experiment shows, the adaptability teaching system based on Bayesian network can offer the network teaching resource of adaptability, is helpful to realize the adaptability learning of network teaching platform.