by YaoweiSun
ABSTRACT
To increase engagement, cultural product sales, and personalization, digital museology must understand visitor behavior. Museums must use data to optimize visitor interaction and cultural merchandising to balance educational mission and operational sustainability. It predicts museum “superstar visitors” product purchases using association rule mining. This research used transactional data from a large metropolitan museum with digital tracking infrastructure to create a behavior-based cultural product recommendation model. From over 50,000 visitor sessions, RFID (Radio Frequency Identification) path tracking, point-of-sale transactions, mobile app usage, exhibit interaction logs, and demographic tags were collected. Data preprocessing included formatting raw behavior into binary transaction records, discretizing continuous variables like time spent and purchase value, and segmenting visitors by frequency and engagement. The study used Apriori and FP-Growth algorithms to find strong association rules linking actions like visiting Exhibit A, interacting with AR features, or attending paid events to book, replica, and digital media purchases. Superstar visitors made more frequent, consistent, and high-lift rules than general visitors, indicating predictive behavior. Events, weekends, and peak seasons improved rule performance, demonstrating how temporal dynamics affect visitor engagement. Behavior-aware personalization may boost museum visitor loyalty and spending. The study shows how behavioral analytics and recommendation systems can make static exhibitions visitor-centered. It highlights superstar visitors’ impact on education and economic outcomes and provides a scalable way for museums to innovate in culture, technology, and audience insight.
