N-mode network approach for socio-semantic analysis of scientific publications

The sciences develop as conglomerates of ideas, texts, and agents. In this study, we propose a n-mode network approach to integrate the network matrices containing social, semantic, and epistemic attributes analytically into a single network and visualization. For example, authors, words (e.g., title words and keywords), and knowledge claims can be attributed to publications as units of analysis. This results in a 2-mode document/attributes matrix which stores both the dimensionality in the data and the interaction terms. Multiplication of this matrix by its transposed provides an affiliations matrix which can be decomposed in specific combinations such as socio-semantic networks. The methodology is illustrated by using the Brugada Syndrome as a specialty in the medical sciences. The proposed n-mode network approach can be extended beyond scientific data; a dynamic extension would allow us to model co-evolutions in more than two dimensions. The proposed approach can be applied to a wide array of studies, such as socio-semantic network analysis or heterogenous networks in actor-network theory.