The Research And Application Of Multi-Agent Technology In The Network Education

The rapid development of Internet and WWW has provided new methods and ideas for long-distance education. Network education, with the network as the vehicle for the transmission of information, is a new teaching method which is a combination of the network and multi-media technology. It has overcome the problem of unidirectionality in the traditional teaching methods, and it has caused the change from a teacher-centered inculcation to a students-centered active learning. And a necessary trend for the future of education it is sure to become. But presently most of the web-based teaching systems are but digitalized versions of the original teaching material, lacking in interaction and insufficient in intellect, therefore making it difficult to practice teaching according to each learner's unique condition. And it is quite possible that the students may go astray from their original target of study. In view of this, to make net education more intelligentized is an urgent problem to be solved and the newly fledgling technology Agent in the area of artificial intelligence has provided a good opportunity to solve the problem.With the application of Agent, intellignetized tutoring on the internet, unification of educational resources of different areas, the creation of virtual classrooms and libraries may be effected, and hence net teaching made more interesting and human, and teaching quality bettered. In recent years, scholars both at home and abroad have carried out researches extensively in intelligentization of net education, for example, put forward by Claude Frasson of University, multi-Agent learning model given by Evandro and Perkusic and Adele's teaching Agent raised by W. Lewis Johnson of South California University, etc. Having the outcome of previous investigations as basis, we comes up with the idea of combining Agent with net education. We have proposed our own net education model, which is based on Multi-Agent. It combined the intelligence of Agent and the constructivist learning theory, and created an individualized and well coordinated net education system. This thesis explicates the functional structure of various Agents, including the Agent of teacher, students, management, character analysis, etc. Besides, there is also a detailed analysis of the implementation of teaching methods and individualization. Agent-based net education pays more attention to the individual condition of each student, as opposed to the traditional teachings. Now, every student can get teaching materials that are suited to his or her own specific conditions, and he or she may have a more individualized learning environment. At the same time, various Agents in the system are entities at once independent of and connected with each other, and this has enhanced the maintainability and expandability of the whole teaching system.This thesis makes an primary explorer and study of the application of multi-Agent in then network education, and makes the base for the deep study in this field.