A method supporting the scientometric use of the bibliographic data sources of HAS

Abstract

In recent years Hungarian scholarly databases, especially those of the Hungarian Academy of Sciences (HAS), have been playing an increasing role in research evaluation. The databases MTA KPA (Publication Database for the Public Body of HAS) and MTA KTA (Database of the Public Body of HAS) are, in conjunction with other data sources, becoming used and referred to in scientometric comparisons. These evaluations, however, usually pose the question: what is an adequate basis for evaluative comparisons, which are the comparable communities of actors (members of the Public Body), supposed to represent a given scholarly field. The study presents an integrated bibliometric classification method for identifying the fields and the corresponding communities that allow valid comparisons. The aim is to classify authors according to their publication record into classes (similarity clusters) that are being set up empirically from bibliographic data retrieved form the MTA KPA. The method is a combination of bibliometric and text mining clustering techniques; (1) it operates based on the information provided by these databases, and (2) it overcomes specific deficiencies of other popular methods, such as bibliographic coupling or co-authorship analysis. The concept is illustrated by a case study, in which a predefined community of the Public Body of HAS, namely a sample of the VIII. Section of Biological Sciences, is subjected to analysis, in order to empirically explore its detailed disciplinary structure. The results are well-interpretable and intelligible as the sub-communities yielded by our method reflect the established taxonomies of biology and the life sciences.