Bibliometric cartography of data science: a large-scale analysis on knowledge integration and diffusion

Bibliometric cartography of data science: a large-scale analysis on knowledge integration and diffusion

연구 분야: Software Development

학회: EPJ Data Science

As widely discussed, data science has been explored and applied in many scenarios. However, little has been known about its knowledge integration and diffusion patterns, particularly its interdisciplinary dynamics. Here, we adopt a citation-based strategy to define the scope of data science and utilize bibliometric methods to map its “cartography” from the perspectives of publications and scholars. Our analysis reveals that, over the past four decades, data science has increasingly integrated a diverse range of knowledge sources, albeit with a pronounced dependence on a limited number of dominant fields. Its knowledge diffusion extends across an expanding spectrum of disciplines, exerting significant cross-domain influence. The diffusion process is driven by both viral and broadcasting adopters, with the former facilitating early-stage dissemination, although a “damping” effect emerges as dissemination efficiency gradually declines. Notably, the interdisciplinary nature of data science is twofold: it not only synthesizes technical innovations but also permeates multiple applied domains, where its application-driven orientation is reflected in domain-specific methodologies. Additionally, we identify the emergence of governance attributes, an increasingly salient characteristic of data science that warrants greater attention. These findings underscore the dual role of data science as both a catalyst for interdisciplinary convergence and a practice-oriented discipline deeply embedded in real-world challenges. Ultimately, data science fosters rapid innovation and facilitates broad knowledge dissemination across academic and practical domains.