Title:	PROPOSAL OF A MACHINE LEARNING MODEL FOR ESTIMATING RESPECTIVE STUDENTS' PROGRAMMING LOGICS IN AN EXERCISE
Author(s):	Ryo Onuma, Sho Onami, Jotaro Kumagai, Hiroki Nakayam, Youzou Miyadera and Shoichi Nakamura
ISBN:	978-989-8704-62
Editors:	Paula Miranda and Pedro Isaías
Year:	2024
Edition:	Single
Keywords:	Programming Logic, Machine Learning, Programming Exercises, Source Code Classification
Type:	Full
First Page:	309
Last Page:	316
Language:	English
Cover:
Full Contents:	click to dowload
Paper Abstract:	Opportunities for programming exercises have expanded in educational institutions such as universities. Accordingly, there is a need to enrich the guidance provided in such exercises. While it is important but difficult to grasp each learner's situation in order to provide accurate guidance, it can also be quite difficult since only a small number of teachers often teach a large number of learners. Moreover, the best way to construct a program (called programming logic) when working on challenges will differ depending on the learner, which makes it even more difficult for teachers to understand the program logic of respective learners. In this research, we aim to develop a model that automatically estimates the programming logic of each learner using machine learning, and then develop a system that presents groups of learners who share the same logic together with corresponding source codes in a way that is easy to understand. This system enables teachers to better understand the unique situations of respective learners. This paper mainly describes an overview of estimating programming logic. Moreover, we describe the experiments on estimation of programing logics, and discuss the basic effectiveness of our model on the basis of the results.
	Go Back