Activity Diagrams as Cognitive Bridges: Revolutionizing Programming Education with AI

In the contemporary digital era, computational thinking (CT) has emerged as a pivotal skill across various disciplines, indispensable for navigating through the intricacies of the information age. Given the incorporation of programming education into K-12 curricula, there is a pressing need for innovative tools to foster CT effectively. This research presents myChatCT, an AI-driven educational assistant, designed to augment high school students' CT and coding competencies by emphasizing problem-solving rather than syntax memorization. Utilizing the ChatGPT API, myChatCT navigates users through the stages of CT—decomposition, pattern recognition, abstraction, and algorithm design—while converting natural language inputs into structured activity diagrams. These diagrams serve as a conduit between conceptual understanding and code execution, providing visual insights into problem-solving strategies and generating annotated code to bolster learning. To assess its efficacy, load tests were conducted to simulate diverse user scenarios, analyzing response times, error rates, and resource utilization. Preliminary findings suggest promising results for enhancing CT and coding proficiency, with areas for improvement identified in system efficiency. Ultimately, myChatCT strives to empower both learners and educators by offering a systematic methodology for cultivating crucial CT skills and revolutionizing programming education.