A Study on Data Analysis of Libraries in United States Using Artificial Intelligence

This study investigates borrowing patterns, usage behavior, and circulation dynamics in U.S. libraries by applying Artificial Intelligence (AI)–driven analytical methods. Using machine learning models, clustering techniques, and regression analysis, the research identifies key factors influencing book popularity and borrowing frequency. Visualizations such as boxplots, scatter plots, correlation heatmaps, and predictive modeling provide detailed insights into how variables like event count, publication year, genre, and purchase count relate to book circulation. Results indicate that event_count is the strongest determinant of borrow_count, while publication year and purchase patterns contribute minimally. Classification models, particularly Logistic Regression, demonstrate exceptionally high accuracy (99%) in predicting book popularity, outperforming Decision Trees. K-Means clustering reveals three distinct usage groups—high, medium, and low circulation clusters—highlighting clear disparities in borrowing behavior. Overall, the study emphasizes the critical role of AI in understanding library usage, enhancing decision-making, and supporting data-driven management of library collections.