Pengembangan Model Pembelajaran Berbasis Kecerdasan Buatan untuk Penilaian Esai Reflektif dan Pemberian Umpan Balik Adaptif bagi Mahasiswa PPG

Abstract

Reflective essay assessment in the Teacher Professional Education Program (PPG) has an important role in measuring students' depth of understanding, reflective attitude, and professional growth. However, time constraints and the subjectivity of assessment are often obstacles in the evaluation process and providing feedback. This research aims to develop an artificial intelligence (AI)-based learning model to support reflective essay automatic assessment and provide adaptive feedback for PPG students. The development approach used is the ADDIE (Analysis, Design, Development, Implementation, Evaluation) model, but it is focused on the D (Development) stage only. The analysis stage was carried out through a literature study and analysis of the needs of PPG students and lecturers related to essay assessment. The design stage develops an AI system framework that is integrated with reflective assessment rubrics and personalized feedback mechanisms. The development stage produced a prototype of a Natural Language Processing (NLP)-based system that is able to analyze the content of the essay and generate automated data-driven feedback. The results of internal trials show that this model has the potential to improve assessment efficiency, provide faster and more relevant feedback, and encourage students' critical reflection in an ongoing manner. This research contributes to AI-based learning innovations in teacher professional education, especially in improving the quality of formative evaluation based on reflective writing.