Community Detection Methods in Library's Books and Borrowers Social Network Segmentation

In this paper, we discuss the application of community detection methods to book- borrowers networks in libraries. The aim is to obtain a segment of books and borrowers that are closely linked to the lending network in the library. This study applies six community detection methods, namely Louvain, Spinglas, Walktrap, Infomap, Label Propagation Algorithm (LPA), and Greedy to identify groups of books and borrowers. Meanwhile, evaluating the effectiveness of this method uses the modularity, performance, coverage, density, community size, and community fit metrics. The results showed that the community detection method was effective in identifying book segments and related borrowers in the library lending network. The Louvain method was found to be most effective in identifying communities with higher quality and better interpretation. The results of segmentation of books and borrowers can support improving library collection management and increasing demand for books, provide insight into patterns of borrowing books to improve library services and user satisfaction.