The Use of Data Mining in Long Tail Service of Library

The research applies data mining techniques to library long-tail services, aiming to enhance the efficiency and effectiveness of library services. A system is established for this purpose, which utilizes a twice-clustering-based text outlier algorithm. This algorithm is designed to identify outliers in the data, which are then analyzed for their potential value in providing new directions for library services. The system is verified using the Wanfang database, a comprehensive Chinese academic database. The verification process involves applying the algorithm to the data and assessing the results. The algorithm successfully identifies a valuable outlier cluster, which consists of three theses. These theses have a high degree of similarity amongst themselves but a low degree of similarity with other theses in the database. This research uncovers a unique outlier cluster, offering a novel direction for library services. These theses represent an underexplored research area, embodying the 'long tail' concept in library services. The study suggests that data mining can identify new service areas, highlighting its potential in enhancing library service provision and development.