Veterinary Medicine Corpus (VMC)

These articles were automatically labeled and sorted using Digital Object Identifiers (DOIs). The PLOS API employed a machine-aided term-mapping algorithm to suggest categories for the published papers, utilizing the PLOS thesaurus (GitHub, 2022). However, the categorization produced by the PLOS website was not optimal.

The default categorization provided by the PLOS thesaurus resulted in an uneven distribution across categories rather than a flat one. Additionally, the suggested categories in Table 2 were generated by machines and based on lexis rather than scientific criteria. To assess the statistical outcomes, we conducted a mock analysis using LancsBox, which revealed high dispersion values (as indicated by the coefficient of variation, CV) for frequently occurring tokens.

This raised concerns about the reliability of developing a discipline-specific word list. As a solution, we opted to refine the categorization process by seeking expert advice. Consequently, we established 4 parent categories and 17 child categories to streamline the classification, as illustrated below: 1.

Pre-clinic (n=352) Disease control and prevention/epidemiology Veterinary microbiology Veterinary parasitology Veterinary pathology Veterinary pharmaceutics and pharmacology Veterinary virology 2. Internal medicine (n=426) Veterinarians Veterinary biometry Veterinary diagnostics Veterinary hospitals Veterinary medicine. Veterinary surgery 3.

Zootechnics (n=121) Livestock care and wildlife sciences 4. Veterinary diseases (n=550) Bacterial diseases Parasitic diseases Viral diseases Other (inherited/genetic diseases/disorders, genetical studies of species regarding diseases) The filnal VMC is 1,449 artcles with 7,962,021 tokens.