Research on application of data mining based on FP-growth algorithm for digital library

Abstract

In the field of the library, there is a huge amount of data from readers' borrowing records every day. We can find an interesting network from the relationship of books by doing data mining works on these records, especially on the analysis of association rules, which can help us to find the needs of readers more clearly. FP-growth (frequent pattern growth) uses an extended prefix-tree (FP-tree) structure to store the database in a compressed form. FP-growth adopts a divide-and-conquer approach to decompose both the mining tasks and the databases. In this paper, we use FP-growth algorithm to analyze the association rule of library circulation records. The results can make great sense to help to improve the quality of library collections.