Learning to program in C remains a significant challenge for first-year students in engineering degrees, often due to abstract concepts and limited personalized support. While generative Artficial Intelligence (GenAI) offers new opportunities for educational tools, most existing solutions focus on languages like Python, leaving C underexplored despite its foundational role in computer science curricula. This study presents the design, development, and comparative analysis of two GenAI-based solutions aimed at supporting C programming learning: a chatbot prototype built on Botpress AI and a custom web platform named BugC. Developed through an iterative process, the first solution enabled rapid prototyping and conversational flow experimentation, while the second offered a fully integrated environment combining theoretical content, interactive exercises, and contextual AI support. A comparative analysis highlights key differences in scalability, customization, maintenance, and pedagogical integration. Results show that while no-code platforms facilitate initial development, they present limitations in cost, flexibility, and long-term sustainability. In contrast, the custom web architecture of BugC offers greater control, adaptability, and alignment with educational needs.

错误:搜索内容不能为空,请输入英文关键词
错误:关键词超出字数限制,请精简
高级检索

BugC: A Conversational GenAI Tutor for Learning to Code

  • María Ferrándiz-Díaz,
  • Andrea E. Cotino-Arbelo,
  • Jezabel Molina-Gil,
  • Carina S. González-González

摘要

Learning to program in C remains a significant challenge for first-year students in engineering degrees, often due to abstract concepts and limited personalized support. While generative Artficial Intelligence (GenAI) offers new opportunities for educational tools, most existing solutions focus on languages like Python, leaving C underexplored despite its foundational role in computer science curricula. This study presents the design, development, and comparative analysis of two GenAI-based solutions aimed at supporting C programming learning: a chatbot prototype built on Botpress AI and a custom web platform named BugC. Developed through an iterative process, the first solution enabled rapid prototyping and conversational flow experimentation, while the second offered a fully integrated environment combining theoretical content, interactive exercises, and contextual AI support. A comparative analysis highlights key differences in scalability, customization, maintenance, and pedagogical integration. Results show that while no-code platforms facilitate initial development, they present limitations in cost, flexibility, and long-term sustainability. In contrast, the custom web architecture of BugC offers greater control, adaptability, and alignment with educational needs.