This article analyzes the conceptual structure of the field of study of innovation systems through the establishment of subfields of knowledge using text mining. The data was sourced from the Web of Science in a search through Boolean operators under the concept “innovation systems” obtaining 6,706 articles in august 2020. For the analysis, a dictionary of categories has been constructed that agglomerate conceptually close families of words comprising keywords, theoretical and geographical approaches. Additionally, the QDA Miner & Wordstat software has been used to exploit the data. As results, it is possible to identify the central role of the concepts of triple helix and open innovation in the field of study of innovation systems; in the same way, it is possible to identify that the concepts of quadruple and quintuple helix related to the greater relevance of social and environmental issues on the agenda emerge strongly. And although the most discussed issues correspond to knowledge transfer, the science, and the economic development; the hottest issues result from the dynamics of innovation systems, its performance, and collaboration actions. Socio-technological changes in the innovation systems, mainly in North America, become less relevant as research interest, while new discussions arise in Europe and Africa on social and environmental sustainability issues. Public policies for innovation systems, management associated with new business models, and sustainability are topics of greatest interest for research. In this way the conceptual structure of the field and subfields of study of innovation systems is explained from the different visualizations through word clusters, the graphic representation of matrices and dendrograms.