Adopting Data Mining Techniques on the Recommendations of Library Collections

In this research, the researchers explored not only the cluster of the readers with similar characteristics, but also the connection between the readers and the book collections of the library by using Data Mining techniques. By doing this, the library will be able to improve the interaction with its readers, and further increase the usage of library collections. The Modified Attribute-Oriented Induction (MAOI) method was introduced to deal with the multi-valued attribute table and further sort the readers into different clusters. Instead of using concept hierarchy and concept trees, MAOI method implemented the concept climbing and generalization of multi-valued attribute table with Boolean Algebra and modified Karnaugh Map, and described the clusters with concept description. On the other hand, the Chinese books in the library collections were classified into four groups with New Classification Science for Chinese Libraries (CCL). Not only the attributes of readers, but also the attributes of library collections borrowed by readers are included in the multi-valued attribute table. After the completion of induction, the reading preferences of the readers with the same characteristics can be learned.