Search

Like other scientific disciplines, sociology can be observed as a social system made up of researchers, institutions, journals and publishers. These relationships are established via conceptual communications which form networks that establish the way in which disciplined are organized. At present, the big data phenomena offers the capacity to use large data collections to analyse social processes. Big scholarly data sources offer sociology immense quantities of data useful to describe and study the evolution of scientific disciplines in detail. In this article we characterised the last thirty years of sociology through its publications in impact factor journals. To do this, we use data about the sociology journals from Journal Citation reports augmented with article information from Microsoft Academic Graph. The analysis starts by describing the journals, countries of origin, languages, publishers, the decades in which they appeared and their impact factor. After this, we evaluate the evolution of numbers of articles and citations as well as co-authorship and gender proportion. Subsequently, we establish four groups of journal types and study their differences in the previous dimensions using hypothesis tests. Finally, we represent the relationships between authors and journals using an affiliation network that allows us to detect groups of journals that form interesting thematic and geographic communities. The novelty of the work consists in having used a data source of the so called big scholarly data with more than 300 million publications. The paper also provides several strategies to select the data of interest among the millions of publications to reduce their dimensionality in order to represent them in the form of a network. The results show a discipline dominated by Anglo-Saxon countries and large publishing conglomerates. The most prominent journals dominate citations whilst methodological journals have a higher degree of co-authorship and thematic journals have the lowest gender bias. The affiliation network between authors and journals contains two large groups, one formed by the pioneering American journals together with quantitative methodological journals and another one made up of English and qualitative methodological journals. [ABSTRACT FROM AUTHOR]