Hierarchical Data Processing and Relational Verticalization for Academic Network Analysis: Transforming CAPES Postgraduate Program Data into Gephi-Compatible Edge Lists

This article presents a formal computational specification for the hierarchical ordering and conditional relational verticalization of datasets extracted from the Brazilian Coordination for the Improvement of Higher Education Personnel (CAPES) Sucupira Platform, with focus on the 2023 quadrennial evaluation cycle. The proposed pipeline converts flat tabular records of postgraduate programs (PPGs), faculty affiliations, productivity indicators, and institutional attributes into a directed or reflexive edge list directly ingestible by Gephi for social network analysis. Through iterative hierarchical sorting and block-transition detection, the method models inter-program collaborations, faculty mobility patterns, and regional asymmetries in Brazilian graduate education. A rigorous, language-agnostic specification is provided — including transition matrices, conditional decision logic, and string construction rules — enabling faithful reproduction in any modern data-processing environment (R, Python/pandas, Polars, Julia, etc.). The framework supports scienciometric investigations into knowledge flows, institutional clusters, and policy-relevant disparities in high-impact fields such as Exact Sciences, Materials, and Environmental Sciences.