The Science Space of Artificial Intelligence Knowledge Production

Abstract

The present contribution seeks to map the development of Artificial intelligence (AI) knowledge as indicated through scientific publications. A set of principal keywords is employed in order to identify those publications that are related to AI in the Web of Science and the metadata provided by the database is then utilised to map the evolution of the field across different scales, such as the national, country, and regional level. This analysis allows for the identification of potential AI “hotspots”, while also establishing places that have been leading in the development of AI knowledge from the onset vis-à-vis those that have managed to catch up over time. In addition, a network that illustrates international collaborative efforts in AI knowledge creation via co-authorships across nations and via the evolution of keywords’ co-occurrence across three decades is illustrated. It is evident that these networks have become denser with time, and that they have changed across regions. The objective of the present analysis is to enhance our understanding about where and how scientific AI knowledge is created, which in turn should encourage and assist future research efforts looking to study AI knowledge and its consequences.