Personalized Learning Path Planning and Course Recommendation Systems Optimized by Genetic Algorithms
TL;DR
The findings show that the SKG-GA methods perform better than multimodal baseline frameworks, achieving important results, with metrics ranging from 90 to 95%.
Abstract
This research proposes a personalized learning path planning and course recommendation system optimized using a genetic algorithm. The system represents learning resources and prerequisite relationships in a structured learning map, encoding each candidate path as a chromosome. Experiments were conducted on a real-world dataset from an online learning platform. By integrating historical and real-time learning data, the system dynamically adapts recommendations to changing learner needs. Data were preprocessed by handling missing values and min-max normalization. Principal Component Analysis (PCA) was applied to reduce dimensionality and extract essential learner and course features. The Scalable Knowledge-Guided Genetic Algorithm (SKG-GA) model is applied to find the most optimal personalized learning path for each learner from a very large number of possible course combinations. Implemented in Python, the findings show that the SKG-GA methods perform better than multimodal baseline frameworks, achieving important results, with metrics ranging from 90 to 95%.