Three IMFD researchers receive Fondecyt Initiation grants

Matías Toro (DCC U. Chile), Sebastián Ferrada (Data & Artificial Intelligence FCFM University of Chile), and Naim Bro (UAI School of Government) are the three IMFD researchers receiving these funds from ANID. The Fondecyt Initiation grants seek to strengthen the development of scientific and technological research excellence by promoting new researchers through the financing of 2- to 3-year research projects in all areas of knowledge.

Differential privacy 

Security and privacy guarantees in computer science are currently one of the most important principles in IT security. In the project by IMFD researcher and academic at the Department of Computer Science at the University of Chile, Matías Toro Ipinza,  "Gradual Differential Privacy: Theory and Applications,", seeks to contribute to the way developers program to ensure privacy and security. 

"The goal is to design programming languages that allow for the gradual analysis of differential privacy, data sensitivity, and security (confidentiality/integrity), combining static and dynamic methods," the researcher points out. In addition to contributing to theory, there are plans to implement practical prototypes that will make it easier for programmers to integrate these tools into their projects.

In the medium term, the research seeks to theoretically develop practical and accessible programming languages that allow developers to incorporate differential privacy and security analysis in a progressive and flexible manner. progressively and flexibly into their projects. This includes the creation of functional prototypes and their validation, laying the foundations for tools that balance precision, flexibility, and ease of use.

Divisions among the Chilean elite

The project entitled "The Urban Elite Divide: A Geospatial Examination of the Brahmins vs. Merchants Hypothesis in Santiago, Chile", by Naim Bro, an academic at the Adolfo Ibáñez University School of Government, explores how the Chilean elite is divided into two groups: those whose influence stems from educational and intellectual achievements ("Brahmins") and those whose influence stems from their wealth and economic power ("Merchants"). 

"I hope to provide a new perspective on political divisions in Chile by analyzing how geography and inequality influence the political behavior of urban elites. This project links spatial and electoral data to map and characterize these divisions, shedding light on how privileged groups shape the Chilean political landscape," says the researcher.

The project is closely linked to data science and the work carried out by the Millennium Institute de los Datos (IMFD). It uses advanced geospatial analysis and machine learning techniques to map and characterize political divisions in Santiago, as well as applying natural language processing (NLP) methods to analyze political discourse and party programs.

Based on concepts from Thomas Piketty and adapted to the Chilean context, the study seeks to analyze the geographic distribution of these elites in Santiago, their evolution over time, and their impact on the Chilean political system. Using geospatial and electoral data, it will investigate how the residential and socioeconomic patterns of these elites affect their political preferences and their relationship with political parties.

Tools for tabular databases with graph capabilities

Knowledge graphs have emerged as a powerful new and versatile model for organizing data, offering a number of advantages over other types of organizations: they have flexible schemas, which make it easy to handle the incomplete and changing data of our modern world. However, most of the information in organizations is stored in tabular data. In "Knowledge Graph Extraction from Tabular Sources,", Sebastián Ferrada, professor of Data & Artificial Intelligence at the University of Chile's Faculty of Chemistry, Physics, and Mathematics (FCFM) and researcher at IMFD, seeks to develop tools that allow users to analyze tabular data with the same flexibility and richness offered by knowledge graphs. This includes developing efficient algorithms and data structures to solve tasks such as community detection, optimal path search, and centrality analysis, while maintaining the semantics of the original data and leveraging the strengths of relational systems.

"In the medium term, I hope that the results of this research will facilitate the adoption of hybrid models where users can benefit from interoperability between relational systems and graphs without having to modify their current infrastructure. This will enable richer and more advanced analysis of existing data," says the researcher. "In the long term, I hope that these tools will serve as the basis for a broader transformation in the way organizations and researchers manage, integrate, and analyze their data, promoting interoperability as a key standard for combining multiple sources of information transparently and efficiently," he adds.

A fundamental aspect of this project is its focus on free software, ensuring that the results are accessible to both academia and industry. "This not only broadens the impact of the project, but also encourages its adoption and adaptation in different contexts," the academic points out.