Knowledge Graph Construction from Library and Information Science Journal Articles

Abstract

Knowledge graphs (KGs) emerge as potential tools for information access and resource discovery in a structured format. It facilitates information retrieval, data integration, and semantic reasoning. Considering the rapid growth of literature publications, a high-relevance search is necessary for researchers and practitioners. There are numerous tools for data organization and knowledge extraction. The knowledge graph is one of them, which depicts structured information with nodes and relationships. Library and information science can help build KGs to grasp the intricate relationships between scholarly works, their authors, institutions, and topics. The knowledge graph technology produces more relevant search results, which makes it easy to explore accurately. This paper examines the construction of knowledge graphs from library and Information Science (LIS) journal articles. A systematic approach is followed to extract entities, relationships, and attributes from LIS literature. The accuracy of the constructed knowledge graph is 89.47% (Recall), 94.44% (Precision), and 91.80% (F1 Score). User satisfaction is 85% in rating their satisfaction with the search results, interface usability, and ease of exploring relationships between entities regarding Scopus and Google Scholar. This paper also discusses this research’s potential areas and challenges in enhancing information organization and retrieval in the LIS domain.