Intelligent Textbooks

Presenter: Joon Suh Choi (Vanderbilt)

Intelligent textbooks have become more popular as advancements in Natural Language Processing (NLP) have made human-machine interaction more accessible, powering interactive features that have been shown to improve reading performance (Clinton-Lisell et al., 2023). The advent of large language models (LLMs) will further change the developmental landscape for intelligent textbooks by enabling higher-quality real-time feedback for user responses to these various features. iTELL is a comprehensive framework for the development of such intelligent textbooks comprising a content authoring and management system, support for multiple interactive, read-to-write tasks like summary writing and constructed response items, a scoring API, and a conversational AI that also acts as a reading support agent for strategic think-aloud. This presentation will provide an overview of the theoretical foundations of iTELL as well as an overview of the general framework, highlighting its potential to transform educational experiences.