The implementation of Project-Based Learning (PjBL) in vocational high schools requires authentic descriptive essay assessments, yet manual grading frequently induces severe teacher fatigue and subjectivity. This study aims to analyze teachers' assessment needs and design a conceptual model for automated PjBL essay grading specifically for the Projek IPAS subject. Utilizing the initial phases of the ADDIE research and development framework, this descriptive study employed preliminary teacher interviews, field observations, and literature reviews. The needs analysis revealed a critical demand for an efficient evaluation system to mitigate educators' cognitive load. Consequently, a conceptual architecture integrating Word Count, Similarity Index, and Keyword Matching was designed strictly as an administrative and technical pre-screening mechanism, while the Timestamp parameter was relegated solely to task management. Ultimately, this automated conceptual model is projected to serve as a cognitive offloading mechanism to substantially reduce administrative burdens and potentially mitigate grading bias, empowering science educators to reallocate their energy toward authentic substantive assessment.
Copyrights © 2026