Generative Knowledge Graph Construction of In-Text Citation Data

Abstract

Recent years have seen an increase in the use of the phrase ”knowledge graph” in academic and professional circles, frequently in conjunction with Semantic Web technologies, linked data, massive data analytics, and cloud computing. Though there has been a marked increase in the availability of scholarly information online in recent decades, all scholarly discourse continues to be conducted through the written word. Scholarly knowledge is difficult to process mechanically in this format. In this research, an extensive dataset is used which is composed of 8,700 academic scholar (research papers sentences). The proposed approach consist of multiple steps; data preprocessing, entity extraction, relationship extraction and knowledge graph construction. We propose a more efficient representation of a scalable knowledge graph by instantly extracting the information from corpus of ACL dataset, and we test whether a knowledge graph can be used as an effective application in analyzing and generating knowledge representation from the extracted corpus of research citations.