- 著者
-
Wenqi Zhou
Takeda Yoko
Song Liu
Urano Yoshiyori
- 出版者
- 一般社団法人電子情報通信学会
- 雑誌
- 電子情報通信学会ソサイエティ大会講演論文集
- 巻号頁・発行日
- vol.2007, 2007-08-29
With the mature of information and telecommunication technology, the traditional testing module is under great change. Item Response Theory (IRT) abandoned the old way of evaluation according to students' raw scores, has been paid more and more attentions. How to combine the real-time and data processing advantages of IT technology to achieve IRT more effectively has been becoming one of the new major issues in the language teaching field. This system tries to achieve the IRT by introducing a "perceptron" technology in deciding the item difficulty coefficients during the tests. By utilizing the perceptron module, system automatically adjusts the difficulty coefficient of the questions according to students' answer so to enable students' correct answer rate convergence to a set value. The difficulty factor of the questions students eventually convergence to would be given as the result of the test in the name of knowledge mastery level.