Students’ attitude towards video-based learning: machine learning analysis with rapid software / Abdullah Sani Abd Rahman ... [et al.]

• Main Authors: Abd Rahman, Abdullah Sani, Meutia, Rita, Hamid, Yusnaliza, Abdul Rahman, Rahayu
• Format: Article
• Language: English
• Published: Universiti Teknologi MARA, Perak 2022
• Subjects: Electronic Computers. Computer Science; Computer software; Application program interfaces
• Online Access: https://ir.uitm.edu.my/id/eprint/74902/2/74902.pdf; https://ir.uitm.edu.my/id/eprint/74902/; https://mijuitm.com.my/view-articles/
• Institution: Universiti Teknologi Mara

The recent computer and Internet technologies have dramatically impacted many facets of education. There has been a rapid rise mainly since the COVID19 pandemic in the use of video-based learning implemented via online classroom setting. Regardless of its usefulness and practicality, the educational technologies adoption has its challenges faced by educators. The main challenge is to get intuition of the students’ attitude that will influence the students’ performances. In supporting the intervention approaches, machine learning techniques have been widely utilized. Another challenge is the difficulty to implement machine learning analysis by the educators. The emergence of rapid software platform can be useful for them, but the existence of these software is unrenowned. The goals of this paper are to: 1) provide fundamental experimental works of the machine learning implementation based on a new rapid software framework and 2) present the ability of machine learning in classifying students’ attitude towards video-based learning. Data were collected from a university level accounting course (n=103), involving students who have different experienced or exposure on video-based online learning. Three machine learning algorithms (Support Vector Machine, Random Forest, and Decision Tree) have been tested on the dataset in a rapid software platform. The results show that all the three machine learning algorithms produced high accuracy (above 95%) prediction results based on the hold-out testing dataset. Furthermore, considering inputs of students perceive on the useful and ease of use of video-based learning as well as excluding the demography attributes in the machine learning models seems more useful in the tested case. This paper presents the fundamental design and easy implementation of machine learning for education domain useful for the inexpert data scientists in many fields.