Network Analysis of Research Base Papers: Metrics and Potential use

This paper describes a common challenge faced by researchers who struggle to find good references for their current domain of study due to the exponential increase in research papers being published daily. The proposed solution involves using deep learning techniques with the BERT Model to obtain a summary of the published paper and its citation papers. Science mapping approaches are then applied to ascertain the similarity between the reference papers and the original paper and rank the citations based on their similarity. Various bibliometric analyses such as citation analysis, co-citation analysis, bibliometric coupling, and network metrics are used to further rank the citations and identify unique citations with citation count. Additionally, author suggestions are provided. The end goal is to develop a visualization tool in the form of a dashboard to help researchers focus on relevant papers.