Geocentrality: Quantifying the Territorial Focus of Scientific Discourse

Abstract

This paper introduces and evaluates the concept of geocentrality, defined as the degree to which a scientific document places geographical territory at the center of its thematic, methodological, or analytical structure. Using a pilot dataset of documents annotated by four independent reviewers, we analyze inter-annotator agreement and validate an automatic geocentrality metric. Results show moderate to good agreement across annotators and strong correlations between human labels and the automated geocentrality score. Based on error analysis, we propose revised decision thresholds and an improved annotation protocol. These findings provide an empirical foundation for computational approaches to territorial centrality in text, with applications to spatial information retrieval and geographically grounded document exploration.

Manuel Morales Almora

Master’s Student in Geospatial Information Sciences

Statistician with experience in data science, spatial analysis, and natural language processing (NLP), currently pursuing a Master’s degree in Geospatial Information Sciences at CentroGeo. His work focuses on integrating statistical methods, geospatial analysis, and artificial intelligence to transform large volumes of data into useful information for scientific research and decision-making.

Alejandro Molina Villegas

Languages, Territory and AI Group Lead

Cátedras Conacyt researcher at the Centro de Investigación en Ciencias de Información Geoespacial (CentroGeo), Yucatán. His research lines include natural language processing, data mining, geoparsing, machine learning, and technologies for the Maya language.