Research on Intelligent Prediction of Learning Styles Based on Multiple Algorithms

With the rapid development of emerging educational and teaching methods such as Massive Open Online Courses (MOOC) in recent years, a large amount of learners' learning behaviors can be recorded and analyzed by the system, laying an important foundation for personalized learning. Based on the Felder-Silverman learning style model, this study introduces intelligent analysis algorithms to dynamically analyze and identify learners' learning styles. It recognizes the learning behavior characteristics of learners through their online learning behaviors, community interaction behaviors, learning content browsing behaviors, click-and-drag behaviors, and other learning processes. The C4.5 decision tree algorithm and Recurrent Neural Network (RNN) algorithm are used to predict their possible learning styles and preferences for learning content formats, aiming to provide learners with learning content and paths that are more suitable for their learning styles, optimize the learning experience, and provide technical support for large-scale, personalized, and high-quality next-generation learning platforms.