The article attempts to summarize the most significant approaches to understanding the digitalization of science. We proceed from the assumption that the development of theoretical models for assessing the “digital turn” will not only help to clarify the changes taking place in science, but will also allow us to better understand them, as well as possibly regulate various aspects of digitalization. We can say that by the 2010s at least four categories (or clusters) of approaches to the conceptualization of digitalization have developed, and each category often operates with its own definitions and a separate conceptual framework. This refers to scientometric, economic, information technology (IT) and sociological approaches to understanding the process of digitalization in science. Even a cursory comparison of their specific characteristics allows us to say that all the paradigms listed above have a number of common features and are based on several fundamental premises regarding the trends in the development of science and education, although an assessment of these trends, as well as an emphasis within each approach can differ significantly. We can single out three most large-scale complexes of phenomena that are in the focus of researchers in the field of digitalization of science: this is the formation of a global academic community thanks to digital services (1), the personalization of higher education (2) and the problem of digital inequality (3). The juxtaposition of these processes, thus, significantly changes several important features of science in general, forcing the academic community to raise questions about the definitions and essence of scientific knowledge once again.