• In the 2023 Knowledge Generation Projects call, six proposals led by CITIC research staff have been selected.
• Their goal is to address challenges in health, sustainability, accessibility, and protection against cyber threats.

 A Coruña, August 26, 2024.- The Ministry of Science, Innovation, and Universities has selected six projects led by researchers from the Center for Research in Information and Communication Technologies (CITIC) of the University of A Coruña within the 2023 Knowledge Generation Projects call. They will receive funding amounting to more than one million euros and will be developed over the next three years.

The programs include the development and application of new ICT technologies in artificial intelligence, advanced statistics, and cybersecurity. They aim to address current challenges in various areas, including health and biomedicine, sustainability, accessibility, economic viability, the generation of reliable intelligent systems, and protection against cyber threats.

More sustainable, accessible, and secure AI to improve people’s lives

The development of new statistical methods for the analysis of complex data with applications in health sciences, biomedicine, and materials science is the goal of the project “Statistical Inference Using Flexible Methods for Complex Data: Theory and Applications.” Led by Mario Francisco Fernández and Ricardo Cao Abad, it will address the research of new non-parametric and semi-parametric methods that can significantly improve the quality of life of patients with diseases such as breast cancer and rheumatoid arthritis.

“Language Technologies from a Green Perspective: Low-Resource Domains”, led by Carlos Gómez Rodríguez and Jesús Vilares Ferro, is the title of the second funded project. In this case, the challenge is to adapt Natural Language Processing (NLP) models to environmental sustainability and economic viability criteria without compromising their performance. The research will focus on techniques to exploit low-resource scenarios, contributing to the creation of more sustainable and accessible artificial intelligence.

CITIC researchers Bertha Guijarro Berdiñas and Amparo Alonso Betanzos will be responsible for materializing the concept of “frugal AI” thanks to the project “Frugal Machine Learning: Empowering AI in Resource-Limited Environments for Real-World Challenges”.The program seeks to develop new machine learning algorithms that work effectively in resource-limited scenarios. The research is part of the trend towards frugal AI, which promotes a more sustainable and democratic use of artificial intelligence, also expanding its application possibilities to environments with data, time, and computational capacity constraints.

“Next-Generation Machine Learning Algorithms for Sleep Medical Records Analysis (NEXT-GEN-SOMNUS)” is the project to be carried out by the teams of Eduardo Mosqueira Rey and Diego Álvarez Estévez. The initiative seeks to improve the analysis of sleep medical records using various next-generation machine learning algorithms, based on self-attention mechanisms, the collaboration of human experts in the learning process, and the inclusion of quantum computing. Through these, it aims to develop more efficient and explainable solutions for the diagnosis of sleep disorders.

The innovative project “Improving Cyber Threat Detection Using Large Language Models for Network Protocols” explores the use of large language models (LLM) applied to network protocols to improve the detection of cyber threats and cyber-attacks. Led by Fidel Cacheda Seijo and Víctor Manuel Carneiro Díaz, it focuses on integrating these techniques into Security Information and Event Management (SIEM) systems, improving the accuracy and efficiency in detecting cyber-attacks.

Finally, “Generating Explanations in Hybrid Intelligent Systems to Ensure Reliability”, co-led by José Pedro Cabalar and CITIC researcher José Santos Reyes, aims to achieve reliable intelligent systems through the generation of knowledge-based explanations, oriented towards human understanding and adapted to accountability, responsibility, and transparency criteria. To this end, the formal basis is the knowledge representation and problem-solving paradigm known as Answer Set Programming (ASP), and the construction of hybrid systems incorporating various extensions in reasoning tasks, as well as machine learning based on evolutionary computation, is studied. The project involves both fundamental theoretical research and the development of software prototypes that will be applied in four real scenarios used as proof of concept.