Analysis of Learning Behavior Characteristics and Prediction of Learning Effect for Improving College Students’ Information Literacy Based on Machine Learning

Abstract

In todays rapidly evolving educational landscape, information literacy has become a fundamental skill for college students, enabling them to adapt to the ever-changing demands of society. It is not only a requisite for selfdirected learning and lifelong education but also a vital aspect of problem-solving, critical thinking, and cognitive abilities. This paper delves into the mechanisms of information literacy teaching by analyzing the diverse learning behaviors of college students and predicting their learning outcomes. Utilizing data from 320 Chinese university students, the study employs the Pearson algorithm to uncover the correlation between information thinking characteristics and learning outcomes. Further, it constructs a predictive model using various supervised classification algorithms, including Decision Tree, KNN, Naive Bayes, Neural Net, and Random Forest. The outcomes offer differentiated intervention recommendations and decision-making insights for enhancing information literacy teaching, improving educational quality, and nurturing innovative talents. The paper underscores the significance of information literacy for problemsolving skills and, as an extension, introduces the powerful XGBOOST & Voting Classifier algorithm to enhance predictive accuracy in this context.