Scientific Collaboration Networks at the Centro Universitario De Los Lagos: a Characterization Based on Complex Networks and Topological Metrics

This work addresses the analysis of scientific collaboration networks based on their topological characteristics, with the aim of understanding the evolution and structure of the co-authorship network at the Centro Universitario de los Lagos (CULagos). To this end, a database of scientific publications is built, from which the co-authorship network is generated and topological measures are calculated to characterize the network. The codes were developed in Matlab. For the databases, the public roster of CONACYT beneficiaries was accessed, from which researchers affiliated with CULagos during the 2005-2023 period were identified, specifically those belonging to the División de Estudios de la Biodiversidad e Innovación Tecnológica. For these researchers, all articles indexed in Scopus were downloaded. The project is organized into the following folders: 1.- Disambiguate Database: Contains the code implemented to identify variations in researcher names, ensuring a clean database in which each author appears only under their primary label. 2.- Global Network Analysis: Generates the collaboration network from the clean database and calculates topological measures. Additionally, weak links are removed to identify more robust collaboration groups. 3.- Cumulative Network Analysis: Builds the collaboration network starting from 2005, iteratively adding each subsequent year’s data to analyze the evolution of various node-level topological metrics, taking into account data from previous periods. 4.- Analysis by Temporal Fragmentation of the Network: Generates temporal subnetworks to analyze the behavior of researchers during specific time periods, using a sweeping approach.