Machine Learning-Based Analysis of Learning Behavior Characteristics and Prediction of Information Literacy Performance Among College Students

Abstract

Information literacy is an essential skill for college students in the modern digital era, helping them effectively access, evaluate, and utilize information for academic and professional growth. This study presents a machine learning-based framework to analyze students learning behavior characteristics and predict their information literacy performance. Student behavioral indicators such as knowledge skills, innovation ability, critical thinking, and responsibility are used as input features for prediction. Various machine learning algorithms including Decision Tree, K-Nearest Neighbor, Naive Bayes, Random Forest, and XGBoost are implemented and compared. Experimental results show that ensemble models provide superior predictive performance, with XGBoost achieving the highest accuracy. The proposed system assists educators in identifying weak learners early, enabling personalized interventions and improved teaching strategies. This approach supports data-driven educational decisionmaking and enhances the overall quality of higher education. Index terms - — Machine Learning, Information Literacy, Student Performance Prediction, Learning Behavior Analysis, Random Forest, XGBoost, Educational Data Mining, College Students, Predictive Analytics, Classification Algorithms