Summary: | 碩士 === 臺北市立教育大學 === 數學資訊教育研究所 === 95 === Fraction concept plays an important role in mathematics field. However, learning the fraction concepts is one of the most difficult parts for children. If the teachers can find out their alternative concepts and offer them the right methods to learn fraction concepts, their misconceptions of fraction can be ruled out effectively.
This purpose of the study was to develop a learning diagnosis system based on Bayesian Networks, which can find out students’ alternative concepts and then provide them supplementary learning paths. The study employed 668 sixth graders from six elementary schools to take the test of fraction concepts. First of all, this research adopted the fuzzy Delphi study to construct the weight of fraction concepts. Then the study computed all of the fractions’ probability, built its Bayesian Network, developed a web-based diagnostic system and used the system to diagnose six graders’ fraction misconceptions. The system could analyze the student’s cognitive degree of fraction concepts and offer the student the remedial learning paths as well.
Taking the advantages of the features of Bayesian network, this research can establish the dynamic conditional probability tables based on the divergent number of students based on. It can also make Bayesian Networks be updated accordingly. As a result, it can more accurately diagnose students’ concepts to all relevant changes. So far, this research has got a satisfactory result in diagnosing students’ fraction concepts.
|