Improving the mathematical input efficiency for mathematical e-assessment systems using an intelligent predictive algorithm

2016 Fiscal Year Final Research Report

• Project/Area Number: 26330413
• Research Category: Grant-in-Aid for Scientific Research (C)
• Allocation Type: Multi-year Fund
• Section: 一般
• Research Field: Learning support system
• Research Institution: Mukogawa Women's University
• Principal Investigator: Fukui Tetsuo 武庫川女子大学, 生活環境学部, 教授 (70218890)
• Co-Investigator(Renkei-kenkyūsha): NAKAMURA Katsunori 武庫川女子大学, 生活環境学部, 講師 (10413329) ; ONO Kentarou 武庫川女子大学, 短期大学部・幼児教育学科, 教授 (50268538)
• Research Collaborator: SHIRAI Shizuka ; FURUKAWA Akira ; SUGINO Ryuuzaburou
• Project Period (FY): 2014-04-01 – 2017-03-31
• Keywords: 数式入力インタフェース / 知的学習支援システム / 数学eラーニング / 機械学習 / 理数系IT活用教育 / 知的インタフェース

Outline of Final Research Achievements

This study aimed to address shortcomings for improving the mathematical input efficiency for mathematical e-assessment systems. We have proposed a predictive algorithm in terms of a perceptron machine learning and developed an intelligent mathematical input user interface (UI), named MathTOUCH, by implementing such the algorithm. We examined the prediction accuracy using a dataset of general 4,000 mathematical formulae from five volumes of a mathematics textbook. The results showed that the accuracy of top ten ranking with our proposed algorithms achieved 85.2%. Finally, we conducted two experiments, i.e. a performance survey and an eight-week learning experiment, to evaluate the efficacy of MathTOUCH implementing STACK e-assessment system. As a result, we have shown that MathTOUCH enabled participants to enter mathematical expressions approximately 1.2 to 1.6 times faster than the standard interfaces and had a high level of user satisfaction in regards to mathematics input usability.

Free Research Field

情報工学