Study And Design On E-learning Adaptive Teacher Model

E-learning has broken through the restriction of time and region. Its instructive mode can't only offer abundant teaching resources but provides the learners with a rich and real-time interactive interface. So it is very significant to design a distributed personalized e-learning system with the aid of computer technology.However, some insufficiencies exist in present network tutoring systems, the ideal system has not realized, summarized as points: First, these systems lack the self-adaptivity and interactivity in general, offer the same courseware for learners and ignore their demands and inhabits. The users are required to adapt these systems, so it is difficult to realize personalized instruction. Second, because the use of hyper-text coursewares, the learners can't be guided effectively and deviate easily from the goal of studying. Third, these systems lack the self-perfection mechanism, and after coming into operation several times they cannot achieve self-adjustion of pedagogical strategies unless be done artificially. Finally, the existing online instructive systems are mainly C/S, they are very huge on the client terminator, and resident programs may operate the client terminator at will, the unsafe factors are easily be caused.To solve these problems mentioned above, we construct aE-learning Adaptive Teacher Model (EATM) based on Agent technology of distributed artificial intelligence and the personalized learning theory, which not only posseses the self-adaptive capacity in the course of instruction but can adjusts the pedagogical strategies by machine learning. We achieved JAVA platform running environment of this model.The basic design philosophy of EATM can be summarized as following: on one hand, we are guided by the studying theory of personalization, and construct corresponding student dynamic model to realize the personalized teaching through monitoring human-computer interaction. On the other hand, we introduced machine learning mechanism to adjust pedagogical strategies.The thesis includes three parts:In the first part, we introduce the background knowledge of intelligent tutoring and Multi-Agents systems.In the second part, we construct the conception model of EATM based on MAS where we propose two-layer structure of pedagogical Agent to achieve self-adaptive pedagogical plan, and the teacher Agent is given self-learning mechanism.The implementation way is introduced in the third part. We construct a three-tier Clent/Server model based on Java platform as the work model of EATM. In this work model, we design the mechanism of communication and interaction of MAS and descript the running environment of this system.