Hybrid prefilling-based library recommended reading study for digital reading

With the rapid development of digital technology, digital reading has gradually replaced traditional paper reading and become the main way for readers to obtain information. Digital reading has promoted the digital transformation of libraries, requiring them to not only manage massive amounts of digital content, but also meet the growing personalized needs of users. However, the traditional library recommendation system can not provide users with personalized reading suggestions effectively because of sparse user behavior data and difficult to guarantee the recommendation accuracy. Combining association rules and collaborative filtering, a hybrid recommendation algorithm is proposed and applied to library recommended reading. The experimental results showed that the number of recommended items of the hybrid prefilling algorithm reached 3,207. The average number of recommended items of the hybrid prefilling algorithm reached 8.07 points. In addition, the proposed algorithm had low error rate, high recall rate, and low false positive rate. The lowest error of the hybrid prefilling algorithm was only 3.67%, and the average recall rate and average false positive rate of the hybrid recommendation algorithm were 95.36% and 3.75%, respectively. The practical application results show that this algorithm can reduce the time and number of students querying books. Meanwhile, students have high satisfaction with the proposed algorithm, with a questionnaire score exceeding 8 points. The research not only provides a new solution for library recommendation system, but also lays a foundation for personalized service in digital reading, which is of great significance to promote the digital transformation of libraries.