Deep Learning-Based Arabic Language Learning Reconstruction Model in Developing 21st Century Competencies

Abstract

This study aims to reconstruct a deep learning-based Arabic language learning model to develop students 21st-century competencies. The study is motivated by the prevailing practice of Arabic language instruction, which remains oriented toward the mastery of grammatical rules, vocabulary memorization, and text translation, and therefore has not optimally developed critical thinking, creativity, communication, collaboration, digital literacy, and character. This study employed a qualitative approach using a library research method. Data were obtained from books, scholarly journal articles, and policy documents and analyzed through content analysis involving data reduction, categorization, interpretation, and conceptual synthesis. The findings resulted in the reconstruction of a learning model that integrates the four Arabic language skills istimā', kalām, qirā'ah, and kitābah with 21st-century competencies through the stages of learning orientation, problem exploration, collaboration, reflection, creation, and authentic assessment. This model promotes contextual, participatory, reflective, and student-centered learning. The novelty of this study lies in integrating the deep learning approach, the 21st-century competency framework, and the characteristics of Arabic language learning into a holistic instructional design that is adaptive to digital technological developments and contemporary educational needs.