Educational AI: Automated Computational Thinking Assessment from Visual Programming Artifacts

• Apeksha Bhuekar

  Author

This manuscript presents a refined iteration of Dr. Scratch, an online platform designed to autonomously evaluate Computational Thinking (CT) proficiencies demonstrated within visual coding artifacts. To overcome the constraints inherent in traditional rule-centric evaluation engines, we introduce a scoring architecture grounded in fuzzy logic. This study surveys pertinent historical efforts, outlines the analytical methodology used to parse Scratch programs, and clarifies the computational parameters evaluated when calculating CT metrics from student-generated submissions. By embedding fuzzy inference, our approach yields continuous, transparent evaluations across fundamental CT vectors. An empirical review involving upwards of 250 Scratch submissions indicates that, in contrast to rigid deterministic models, this fuzzy framework delivers superior precision, closer correspondence with teacher appraisals, and enhanced clarity. We outline initial outcomes from this research, propose upcoming trajectories, and tackle prevailing bottlenecks in digital pedagogical evaluation. Ultimately, this research propels Educational AI forward, offering a more sophisticated, instructionally aligned paradigm for measuring computational reasoning.

• PDF

Copyright (c) 2026 International Journal of Intelligent Systems and Data Science

This work is licensed under a Creative Commons Attribution 4.0 International License.