WIP: AI Tutor for Assessing and Training Engineering Students in Reading Comprehension

This WIP research to practice paper observes in past studies on engineering education that they highlight a persistent challenge: Engineering students struggle with reading comprehension despite the need to read and comprehend diverse sources of information from a variety of domains. The existing pedagogies of reading comprehension education lack scalability for assessments, leading to poor instruction. To address this gap, we developed an AI-based system to assess engineering students’ degree of reading comprehension and offer personalized feedback. The AI system is trained to recognize the emotional states of a reader from their physiological changes measured through facial expressions. The central idea behind our research is that the emotional responses of readers are correlated to the ‘sense’ encapsulated in the text. Therefore, the degree of reading comprehension can be assessed by studying the relationship between the ‘sense’ of the text and the AI-detected emotion of the reader. The proposed AI tutor can enable instructors to gain valuable insight into the reading comprehension skills of students using a technological intervention that is more accessible to an engineering mindset.