In order to improve the situation of low intelligentization in ICAI system, the application and developing trend of ICAI is introduced in this paper. We also put forward two advanced methods which are based on teaching cybernetics. By synthesis the knowledge of pedagogics, psychics, artificial intelligence, we have discussed how to solve the problem of the selection of teaching methods in ICAI system, and we tried some new instructive methods to solve this problem theoretically and practically.Firstly, we analyzed all kinds of possible modes on how to select teaching methods, and two of the best modes are brought forward and researched. Secondly, the mode based on the expert system of the uncertainty reasoning in teaching methods is discussed emphasizely, and the theory of certainty factor is applied to the mode, then we develop an expert system program by using CLIPS language. In the system, fuzzy knowledge expression and reasoning based on the production rule are realized, so that we can perferably realize the intelligent and individual property of ICAI. Thirdly, when we select teaching methods in practice, we may find the problem rule deformity and data fuzzy. So it is nature for us to introduce fuzzy neural networks in our mode. We design a fuzzy neural networks model in VC++ language and present the test results in the end of this paper.At last, by compare and contrast the results of two modes, we get the relative merits of them.
|