Research on Constructing Adaptive Learning Paths for University Administrative Staff Based on Large Language Models

This research explores an innovative approach to constructing adaptive learning paths for university administrative staff based on large language models. By analyzing the educational capabilities of large language models, adaptive learning theories, and the learning characteristics of university administrative personnel, the study proposes a comprehensive theoretical model and technical framework. The system employs multidimensional learner modeling, knowledge tracking, and dynamic path generation algorithms to precisely identify learners' knowledge states and recommend personalized learning content. Experimental results demonstrate that, compared to traditional fixed learning paths, this approach significantly reduces learning time (20.0%, p<0.01), improves knowledge mastery (20.3%, p<0.01) and application ability (19.4%, p<0.01), and enhances knowledge retention (31.4%, p<0.001). Different types of learners benefited to varying degrees, with younger learners, visual learners, and those with high technology acceptance showing more significant effects. The research confirms the feasibility and effectiveness of large language models in constructing adaptive learning paths, providing a new technical approach and methodological framework for the intelligent upgrade of university administrative staff training systems, and indicates directions for future research.