Emerging patterns and trends in the international scientific structure of the hate speech domain

Abstract

The aim of this study is to identify the patterns and trends in the international scientific structure on hate speech. It revealed the network of scientific collaboration, the structure of co-citations, the areas of knowledge linked to the topic, and the subjects that define the trends in the domain. We retrieved 441 articles that include the expression hate speech in the title, abstract or keywords field from the Web of Science Core Collection for the period 2009-2018. The collaboration and co-citation networks were modeled using Bibexcel version 2017 and Pajek for the analysis and visualization. Latent Dirichlet Allocation was used as a natural language processing technique for the title, abstract and keywords fields. The results show a significant increase in publications since 2013 and a peak in 2018. The Social Network Analysis showed that despite the incipient presence of hate speech in the scientific literature, there is a prominent theoretical core of publications on the topic which is frequently cited by the international scientific community. There is also a core of works that is recognized as highly relevant in the theoretical, conceptual and methodological approaches to the study of hate speech.