Biblium: An Advanced Python Library for Bibliometric and Scientometric Analysis

This paper presents Biblium, a comprehensive Python library and graphical application for bibliometric and scientometric analysis. With over 200,000 lines of code across over 350 classes and nearly 5,000 functions, Biblium provides an extensive Python-native solution for bibliometric research. While replicating core functionalities of the widely-used R package Bibliometrix, Biblium introduces significant innovations in three key areas: (1) comprehensive group-based comparative analysis with statistical association testing, (2) predictive modeling capabilities for group membership classification, and (3) a full-featured graphical user interface for researchers without programming expertise. The library natively integrates data from major bibliographic databases including Scopus, Web of Science, OpenAlex, and PubMed, with prototypic support for additional sources, enabling analysis workflows from data import through publication-ready exports in multiple formats. Validation testing against five established Python bibliometric libraries (litstudy, metaknowledge, pybibx, pyscisci, and scientopy) demonstrated 100% agreement with consensus values across 30 compared metrics, confirming computational accuracy. Performance benchmarks revealed that Biblium excels in network analysis operations—completing keyword co-occurrence analysis 25–501% faster than competing libraries—while maintaining consistent, predictable execution times across repeated measurements. We demonstrate Biblium's unique capabilities through comparative analysis with existing tools, highlighting its advantages in subgroup analysis—a functionality absent from current Python alternatives and only partially available in R-based solutions. Biblium is openly available on GitHub and PyPI.