Intelligent Tutoring Systems Powered by Generative AI: Advancing Personalized Education and Overcoming Challenges

Brain signals analysis can be applied to students to identify their subject- or topic-specific struggles while they are listening to lecture videos. This re- search work proposes a threshold-based anomaly detection algorithm to identify the topics that cause students the most distress by analysing a dataset available at https://www.kaggle.com/datasets/wanghaohan/confused-eeg/data. Firstly, it identifies the index and timestamp of distress data points of students’ brain signals, extracts the respective portion of transcribed textual content, given as prompt to ChatGPT or Bard via application programming interface (API) calls, get back the answers, and display it in a pop up window to the students. The proposed generative artificial intelligence (Gen AI)-powered intelligent tutoring system (ITS) captures almost 99% of confusion points and source confusion topic in brain signals as it uses context- aware algorithm thereby improving leaners’ clarity on the source confusion topic after reading the pop-up window content generated by Gen AI.