WIP: Adaptive Presentation Training Powered by AI

In recent years, the use of AI technology to evaluate student presentations has become increasingly prevalent among researchers. These evaluations generally focus on key presentation skills such as flow, organization, pronunciation, and the use of sophisticated vocabulary. In addition, they assess the quality of the content, including the accurate use of key terms and the extent of students’ understanding of the subject.In this study, our goal is to push the field forward by utilizing AI technology not only to assess both presentation delivery and content creation, but also to provide a step-by-step guide—functioning as an AI tutor—to help students practice and enhance their presentation skills. For example, if a student’s eye contact is insufficient, the AI tutor will encourage them to focus on a specific point. This AI tutor delivers personalized and adaptive feedback to help students improve their presentation abilities. Furthermore, our platform supports ongoing enhancement of content quality; for instance, if the vocabulary used is not sufficiently professional, the AI tutor will suggest more suitable alternatives. Currently, we employ a general-purpose model, but we intend to integrate domain-specific content analysis and further optimize terminology recommendations in the future. We plan to conduct a survey to gather student feedback on this approach, and we anticipate that the results will demonstrate significant improvements in presentation skills, as students will be able to access the AI tutor anytime and anywhere.