The present work demonstrates the use of graph machine learning techniques to analyze the co-authorship network among affiliated authors of the Alianza EFI Project. The document is divided into three chapters: the first one provides a comprehensive overview of the global and local context of Artificial Intelligence (AI) in a way that justifies the significance of working with AI topics in today’s world. The second chapter is dedicated to constructing the theoretical framework for working with graphs and machine learning. The final chapter showcases the results of implementing graph machine learning for predictive tasks at the node, link, and community levels. Specifically, this chapter reveals that the Alianza EFI project involves contributions from 390 unique authors, associated with 112 distinct institutions, resulting in 274 unique products. It also demonstrates that the Universidad del Rosario plays a central role in institutional collaborations, in contrast to the other institutions within the alliance. Finally, after applying graph machine learning techniques, it was observed that these strategies enable the alliance to identify new research topics for authors, establish new connections among isolated authors, and discover new communities of research interests.
