Design and Development of a Learning Style Identification Application for JPTK Students using the K-Nearest Neighbor

Abstract

Learning styles are crucial for all students, as the chosen learning style can greatly assist them in learning. The data source for this research originates from questionnaire results distributed to JPTK students of the 2019-2021 cohorts, which were used to assess the effectiveness of a learning style product on the students' JPTK website. This study employs the K-Nearest Neighbor approach, which utilizes the principle of nearest neighbors to categorize students' learning styles based on provided features. The data used in this research is derived from the website that students use to input information about their preferred learning styles. Various elements, including visual, auditory, and kinesthetic preferences, are present in the questionnaire on the website. Subsequently, the data is processed and fed into a Python K Nearest Neighbor model to predict students' learning styles and nearest neighbors. The evaluation results indicate that the developed classification model achieves a reasonably high accuracy level of 93%, making it a useful tool for effectively and efficiently identifying students' learning styles. It is hoped that implementing this learning style classification model will benefit the field of education. By understanding students' learning styles, educators can create more tailored lesson plans, enhance learning outcomes, and reduce the likelihood of knowledge loss.