Developing a Model Based on Students’ Perceptions of Using Artificial Intelligence Technologies to Support Self-Directed Learning

Abstract

This study investigated the creation of a model based on students' opinions regarding the use of artificial intelligence (AI) technologies to facilitate self-learning for computer science and educational technology students. The objectives were to gauge students' knowledge of AI, evaluate how they used these tools for self-study, and pinpoint the difficulties and potential developments in the use of AI in higher education. The study used a questionnaire with 45 statements arranged along three primary axes: self-learning applications of AI, awareness of AI and its significance, and future trends and problems in its use. A sample of students from the Faculty of Specific Education provided the data, which was then evaluated using weighted averages, standard deviation, descriptive statistics, and student classification into LOW, MEDIUM, and HIGH categories. To find differences between the axes, ANOVA was also employed. The Random Forest approach was used to create a predictive model. A Calibrated Classifier was used to calibrate the results, and a Label Encoder was used to translate categorical traits into numerical values. This methodology was used to evaluate both individual and group performance and customize learning paths to students' skill levels. The findings demonstrated that students use artificial intelligence's capabilities to improve self-learning, time management, receiving feedback, and reviewing academic performance data, and they have a moderate to high level of understanding of the technology's significance. While student level analyses showed an increase in the HIGH category following system deployment, ANOVA results showed consistency in student assessment across the three dimensions. This demonstrates how well the paradigm enhances self-learning and fosters original thought. Additionally, the model improved the efficacy of self-learning and made it possible to offer tailored instructional advice. Expanding the system's deployment, creating faculty training programs, bolstering digital infrastructure, offering continuing student assistance, and promoting AI-enhanced collaborative learning were among the recommendations. Additionally, recommendations for further research were given, such as increasing the sample size, testing the system in various learning situations, creating sophisticated adaptive learning algorithms, researching collaborative learning, and assessing the system's long-term effects.