Bibliometrics effects of a new paper level classification

This study presents a comparative analysis between two scientific document classification systems. The first system employs the Scopus journal-based assignment method, adapted to a fractional model, while the second system uses an item-by-item system based on reclassified references according to the origin of the citers. The study's results are divided into three different sections: the first involves comparisons at the Scopus area level, the second examines comparisons at the category level, and the third tests various bibliometric indicators to identify the variations between the two systems. Highlighting the characteristics of the paper level system, it offers a reduction in the number of categories to which each document is assigned, achieving higher values of single-category assignment compared to the All Science Journal Classification (ASJC). When reclassifying areas and categories, the paper level system tends to accentuate differences at the extreme values, increasing the size of the largest categories and reducing that of the smallest ones. Moreover, the paper-by-paper system provides more homogeneous distributions in normalised impacts and adjusts values related to excellence more uniformly.