Development of a Formative Assessment Chatbot for a 100-Level Course in Cybersecurity
Clement Amone Keyamo, Peace Ogechi Edun, Abimbola Adebisi Baale · Faculty of Natural and Applied Sciences Journal of Computing and Applications · 2025
This study presents Cypanion: a formative assessment chatbot developed to support first-year undergraduate students in learning foundational cybersecurity concepts through interactive dialogue and real-time feedback. Classroom observations revealed limited prior exposure to cybersecurity among students, often leading to disengagement and poor performance. To address this, Cypanion was built using the DIET-enabled RASA framework for natural language understanding, supported by Python, SQLite3, and the PyCharm API. The system was trained on domain-specific datasets comprising instructional content, question–answer pairs, and conversational tasks including registration, assessment, and formative feedback. Evaluation focused on two natural language understanding subtasks: intent classification and entity extraction. The chatbot achieved 95% precision, recall, and F1-score for intent classification, and 85% precision, 75% recall, and 80% F1-score for entity extraction. These results demonstrate Cypanion’s effectiveness in accurately interpreting student inputs and generating relevant educational responses. The findings suggest that formative assessment chatbots can enhance engagement, support personalized learning, and scale assessment delivery in technical domains like cybersecurity. Future improvements will include speech-based interfaces and exam security modules for deployment in remote learning environments.