A methodology oriented to unstructured data of scientific production for the temporal evaluation of research groups
Scientometrics; Machine Learning; Complex Network Analysis; Graph Embeddings; Research Groups
Funding agencies and research institutions often employ quantitative methods and scientometric techniques to evaluate scientific groups. These evaluations typically rely on a single type of metric, whether it is based on counts (such as the h-index) or derived from Complex Network Analysis. However, the use of multiple measurement approaches and the proper exploration of the temporal dimension of academic production is still a recurrent issue. Particularly, an underexplored approach involves combining these indicators with Machine Learning and Graph Embedding techniques, which could enhance the evaluation process of research groups. In this context, this study proposes a Graph Data Science-oriented methodology to analyze scientific teams over time. Through a case study, the results suggest the feasibility and suitability of the proposed method for quantitatively assessing research groups. The presented approach has the potential to provide strategic and proactive insights for scientific teams, contributing to a better understanding of their dynamics and limitations.