Applications of Natural Language Processing in Adaptive Training

Morgan Ulinski, Victor Hung

Abstract: This paper explores two key applications of natural language processing (NLP) in adaptive training: (1) content generation and (2) simulation-based training. We introduce a content generation tool which leverages retrieval-augmented generation to help instructional system designers create course materials more efficiently. The tool reduces development time while maintaining human oversight to ensure accuracy and pedagogical quality. For simulation-based training, NLP enables automated scenario generation (ASG) and intelligent computer-generated forces (CGFs). ASG processes operational documents to extract mission parameters, reducing manual effort in designing training scenarios. Intelligent CGFs create adaptive, realistic adversaries in simulations, enhancing engagement and skill development. Despite these advancements, challenges remain in ensuring content accuracy, knowledge representation, and adaptive behavior in AI-generated training environments. Addressing these issues requires refining NLP techniques, improving pedagogical alignment, and enhancing validation mechanisms. This paper highlights NLP’s potential in adaptive training and outlines future directions for AI-driven learning systems.

External IDs:dblp:conf/hci/UlinskiH25