A Study on Digits and Mandarin Phonetic Symbols of an Intelligent CAI System Using Neural Networks and Fuzzy Inference Rules

• Main Authors: Ruan Chi Haw, 阮志豪
• Other Authors: 郭耀煌
• Format: Others
• Language: zh-TW
• Published: 1998
• Online Access: http://ndltd.ncl.edu.tw/handle/24474768235643867631

碩士 === 國立台南師範學院 === 資訊教育研究所 === 86 === In this thesis, we apply the neural network techniques and fuzzy set theories to construct an intelligent computer-assisted instruction (CAI) system for teaching elementary students to learn and write the standard digits and Mandarin Phonetic Symbols. The major researches include: information techniques for handwritten characters recognition, artificial intelligence, and multi-media CAI system. The performance of this system was evaluated by the quasi-experimental of education research. The examinee samples for the research are forty first-grade students. Twenty students are used as the control group and the rest students (twenty students) are used as the experiment group. The experimental results are summarized as the followings: 1.The standard degree of written characters is evaluated by the fuzzy inference rules. (1) At the post-testing, the standard-degree of experiment group was higher than the control group. (2) At the post-testing, the standard-degree of experiment group was better than that in the pre-testing (3) The correlation of standard-degree between the output of fuzzy inference rules and the output of neural network are positive and high correlation. 2.The standard-degree of written characters is evaluated by the neural networks. (1) At the post-testing, the standard-degree of experiment group was better than the control group. (2) At the post-testing, the standard-degree of experiment group was better than that in pre-testing stage. Experimental results verify the high value of this research, and the experience of this research is useful to related works.