Using digital-pen handwriting traits to analyze students’ personological and assist teachers to adapt teaching on junior high school mathematical course

• Main Author: 張家榮
• Other Authors: 區國良
• Format: Others
• Language: zh-TW
• Online Access: http://ndltd.ncl.edu.tw/handle/18814240957595500860

碩士 === 國立新竹教育大學 === 應用數學系碩士班 === 96 === Recent researches indicated that there exist relationships between students’ personality characteristic and learning performance. Therefore, teachers need to detect students’ personality characteristic and then promote students to learn in an efficient way. This paper established a system for teachers to detect and analysis students’ personality characteristic with graphology theory and machine learning technology. Two machine learning technologies were employed in this research: decision tree and Bayesian network. The decision tree assists teachers to predict students’ personality by the classification rules. It also illustrates the relationships between handwriting traits and personality graphlicaly. Meanwhile, the Bayesian network assists teachers to explore the causal relationship of handwriting traits and personality graphicaly. Therefore teachers can improve students to learn by considering students’ personality effects. The result illustrates that there were 40 kinds of handwriting traits and 4 kinds of personality characteristics extracted succefully in a short time. Thus, teachers can reduce the efforts to detect student’s personality characteristics by using this system, and then adjust his/her teaching strategy for adating students to learn.