Presentation 2023-01-20
A System for Estimating Adaptive Solution Based on Transition of Each Learner's Source Codes
Soichiro Sato, Yoshiki Sato, Shoichi Nakamura, Youzou Miyadera,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) Analyses of source codes created by learners are often used to support teaching in programming classes. However, most studies aim to estimate the learners' understanding and identify those who need help. The purpose of this study is to develop a system that automatically estimates a solution method (source code) from the transition of source codes submitted by a learner who has had difficulty with the programming tasks in the past. The estimated source code should be easy for the learners to understand, and teachers can use it to provide adaptive solution support. An approach to developing this system is to compile source codes and generate abstract syntax trees and then convert to token columns and weights. These are classified for each solution by clustering and then labeled. After inputting the transition of labeled source codes, a machine learning model is created to output the labeled source code (solution method). In the development of machine learning models, clustering results have a significant effect on estimation accuracy, therefor we compared and examined the clustering threshold value for each task to estimate the solution. The machine learning model created in this study was evaluated using an existing machine learning evaluation method, and the results suggested that our developed model can automatically estimate the solution that is easy for learners to understand with high AUC.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) Programming Learning / Solution Estimation / Machine Learning / Source Code History / Educational Support
Paper # ET2022-57
Date of Issue 2023-01-13 (ET)

Conference Information
Committee ET
Conference Date 2023/1/20(1days)
Place (in Japanese) (See Japanese page)
Place (in English) Hyogo College of Medicine and Online
Topics (in Japanese) (See Japanese page)
Topics (in English) Embodies Knowledge and Skill Education, etc.
Chair Kenji Watanabe(Hiroshimai Univ.)
Vice Chair Hisayoshi Kunimune(Chiba Inst. of Tech.)
Secretary Hisayoshi Kunimune(Mejiro Univ.)
Assistant Kazuaki Yoshihara(Kinki Univ.) / Toru Kano(Tokyo University of Science)

Paper Information
Registration To Technical Committee on Educational Technology
Language JPN
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) A System for Estimating Adaptive Solution Based on Transition of Each Learner's Source Codes
Sub Title (in English)
Keyword(1) Programming Learning
Keyword(2) Solution Estimation
Keyword(3) Machine Learning
Keyword(4) Source Code History
Keyword(5) Educational Support
1st Author's Name Soichiro Sato
1st Author's Affiliation Tokyo Gakugei University(Tokyo Gakugei Univ.)
2nd Author's Name Yoshiki Sato
2nd Author's Affiliation Tokyo Gakugei University(Tokyo Gakugei Univ.)
3rd Author's Name Shoichi Nakamura
3rd Author's Affiliation Fukushima University(Fukushima Univ.)
4th Author's Name Youzou Miyadera
4th Author's Affiliation Tokyo Gakugei University(Tokyo Gakugei Univ.)
Date 2023-01-20
Paper # ET2022-57
Volume (vol) vol.122
Number (no) ET-348
Page pp.pp.5-10(ET),
#Pages 6
Date of Issue 2023-01-13 (ET)