Citation network analysis for journal selection of university library

Abstract

Every academic year, a university library must subscribe tons of journals to provide the best resource for the students and researchers. However, subscribed journals have been used unequally and some relevant journals are not subscribed. Therefore, the library needs supporting information about usedness and demand of journals for efficient collection management. In this paper, citation network analysis is applied to analyze the articles published by university researchers and their references. The result is two overlapping lists of journals that are recommended for the library collection i.e. (a) journals that the university researchers have published in and (b) journals that the university researchers have cited. The importance of the journals is scored based on multiple centrality metrics. We used articles published by Chulalongkorn University (CU) researchers between 2016 and 2018 to analyze and compared the results with articles publishing in 2019 to validate the proposed method. There are two levels of experiment in this research, the first experiment considers articles for the whole CU and second experiment considers top 5 faculties with the most publications. The results for both experiments show that top journals have not changed much over the years. The experimental results show that the ranked journal lists created from publication data of previous years closely matches the list produced from publication data in the subsequent year. The proposed model can help the university library to understand usedness and demand of journals and give suggestion of journals that should be included in the university library collection based on past publication data. The journal lists can also help researchers to choose journals for publication and reference.