Alejandro Beltran is a postdoctoral research associate within the public policy programme. He received his PhD from the University of Arizona where in his dissertation he identified the institutional and political determinants of corruption investigations in subnational audit agencies of Mexico. As a computational social scientist, he uses machine learning and natural language processing to generate quantitative measures of often unmeasurable phenomena from text in Spanish. Alejandro completed his undergraduate studies in Public Policy at the Universidad Autónoma de Sinaloa. Prior to joining the Alan Turing Institute, he was a postdoctoral researcher at the School of Government and Public Policy at the University of Arizona where he developed tools for tracking migrant caravans and gangs across Central America using local newspapers. He has also collaborated with the Inter-American Development Bank in building readily available data on subnational public finances for countries in Latin America.
Publications and Working Papers
Beltran, A. (2023). Fiscal data in text: Information extraction from audit reports using Natural Language Processing. Data & Policy, 5, E7. Paper Code
Osorio, J., & Beltran, A. (2020). Enhancing the Detection of Criminal Organizations in Mexico using ML and NLP. In 2020 International Joint Conference on Neural Networks (IJCNN) (pp. 1-7). IEEE. Paper
Osorio, J., Reyes, A., Beltrán, A., & Ahmadzai, A. (2020). Supervised event coding from text written in Arabic: Introducing hadath. In Proceedings of the Workshop on Automated Extraction of Socio-political Events from News 2020 (pp. 49-56). Paper