Improving College Students' Information Literacy with Machine Learning: Learning Behaviour Analysis and Effect Prediction

Abstract

In addition to being a prerequisite for selflearning and lifelong learning, information literacy is a fundamental skill for college students to adjust to contemporary societal demands. Using the rich and varied characteristics of information literacy learning behavior to perform the learning impact prediction analysis is an efficient method of revealing the information literacy teaching mechanism. By building a prediction model of learning impact based on information literacy learning behavior characteristics, this article examines the features of college students' learning behaviors and investigates the predicted learning effect. 320 college students' information literacy learning data from a Chinese institution were utilized in the project. The learning behavior traits of college students' information literacy are analyzed using the Pearson method, which shows a strong relationship between the traits of information thinking and learning impact. To categorize and forecast the learning impact of college students' information literacy, supervised classification algorithms such Decision Tree, KNN, Naive Bayes, Neural Net, and Random Forest are employed. The Random Forest prediction model is shown to perform best in the learning effect classification prediction. The F1 score is 89.39%, the accuracy is 92.50%, the precision is 84.56%, the recall is 94.81%, and the Kapaa coefficient is 0.859. In order to modify information literacy teaching behavior, enhance the quality of information literacy instruction, optimize educational decision-making, and foster the long-term development of exceptional and creative talent in the information society, this paper proposes differentiated intervention recommendations and management decision-making references in the information literacy teaching process of college students.It was encouraging to do study on the direction and way of thinking for the sustained growth of information literacy training.