Analysis of Scientific Growth Patterns and Citation Distribution Driven by Educational Data Based on ArXiv Database

Research on the number of scientific papers and citations in education data has yielded insights into potential patterns and general laws of scientific growth. We use the arXiv database, encompassing over 1.88 million scientific papers and their citations spanning the years 1991 to 2022. Through an analysis of the growth rate of the number of scientific papers, the findings indicate a gradual slowdown in the scientific growth pattern over different periods. Comparative assessments of growth rates across the disciplines of physics, mathematics and computer science reveal diverse growth patterns in different disciplines. The exploration extends to the analysis of directed citation network constructed by scientific papers, highlighting that only a fewer paper exhibits higher connectivity. This observation emphasizes the significant contributions made by few papers and helps to understand the relationship between collaboration and information dissemination in scientific research. Furthermore, an analysis of the citation distribution and cumulative citation proportion of scientific papers over the preceding ten years. The results show a discernible time decay phenomenon in citation patterns and emphasize the significance of recent research in advancing the scientific process. This find contributes to a deeper understanding of the dynamics of scientific growth, collaboration, and the impact of contemporary research in the academic landscape.