Using LLMs, Knowledge Space Representations and Worked Examples to Support Adaptive Cybertraining in Construction Education

Document Type

Conference Proceeding

Publication Date

1-1-2025

Abstract

Construction education faces a number of challenges as the industry moves to embracing high-technology systems such as building information modeling, robotics, digital twins, AR/VR, and other systems. These systems usually need specific training to take advantage of, but the construction industry often relies on on-the-job or trade schools where there is limited expertise among instructors and limited resources to access technology. We describe a new pilot program for improving cyber-instruction in construction education that aims to improve access and support for students both at the university level and in non-traditional education and on-the-job training contexts. It explores using LLMs and chatbot to help learners explore new materials; knowledge-space representations to understand the necessary prerequisite knowledge (and move past information already mastered), and worked examples to improve student mastery of complex procedural skill. In this paper, we describe the approach of our ongoing project and illustrate some of the advances made.

Publication Title

Lecture Notes in Computer Science

ISBN

[9783031929694]

Link to Full Text

Share

COinS