Development of a semi-automatic bibliometric system for publications on animal health and welfare: a methodological study
Résumé
Bibliometrics is a common research instrument for systemic analysis of scientific
progress in many disciplines, including veterinary sciences. However, bibliometric
analyses are generally biased by limited information included in scientific productions
databases. The aim of this study is to propose a method for implementing databases on
livestock health and welfare in a bibliometric perspective. By using the detailed classification
of the scientific disciplines and the list of keywords of the CAB Abstracts , queries
were first built to combine 120 descriptors related to animal species and 1680 descriptors
related to the given thematics. Then, the extraction was done from the Web of Science .
The overall process focused on the research institutes localised in Europe and on publications
written in English. To assess the database accuracy, supplementary filters were
validated and applied to discard non-specific terms and neighbouring topics. Additional
fields for species groups, infectious and non-infectious diseases, welfare components and
geographical regions were incorporated according to thematic terminologies specifically
established. The final database contains 57,523 articles published over a 12 years’ period
(2003–2014). The developed method, based on the complementary use of both the CAB
Abstracts and Web of Science , can be a reference to allow an adequate semi-automatic
retrieval of relevant publications on livestock health and welfare, which will be of use to
implement bibliometric study to produce a realistic assessment of tendency for driving
science policy. Interestingly, it can be also a model process to be easily applied to
implement bibliometric studies for any other scientific thematic domain.