Abstract

Social work researchers are increasingly required to process and use digital information appropriately. The aim of this scientometric study is to demonstrate how to use big data to analyze social work research questions. The study uses topic modeling to automatically extract topics from 42,001 abstracts retrieved from the Social Work Research Database. Some of these topics are used to illustrate the method’s potential and to shed light on the core semantic structure of disciplinary social work. Some topics, such as child protection and addiction prevention, have become a focus in the social work literature, while other areas have received progressively less attention. Against the background of the outlined problems regarding the interpretation of automatically generated results, analysis methods like topic modeling should be understood as complementary to the more classical research approaches.