News

The CEREBRA-AI project recently held a project meeting in Thessaloniki, bringing partners together to share progress, review ongoing activities, and coordinate next steps toward achieving the project's objectives. The meeting provided a valuable opportunity to strengthen collaboration among the consortium members and ensure a common understanding of the project's technical developments and planned activities.

One of the main topics of the meeting was the presentation of the project's technical architecture, which provided an overview of the system's main components and their interactions. The partners also discussed the different use cases addressed by the project, exploring their requirements, objectives, and expected contributions. These discussions helped align the technical developments with the practical needs and challenges that CEREBRA-AI aims to address.

GSI-UPM is actively contributing to several key areas of the project. In Work Package 3 (WP3), its main contribution focuses on developing the continuous learning engine, a key component that supports the system's ability to adapt and improve over time. In Work Package 5 (WP5), GSI-UPM is involved in research and development activities related to explainability, addressing the need to make the system's processes and outcomes more understandable and transparent. In addition, GSI-UPM actively participates in the project's use cases, contributing its expertise to their definition and development.

The meeting in Thessaloniki marked another important step in CEREBRA-AI's progress, fostering knowledge exchange and coordination between partners and helping consolidate the technical foundations for the next stages of the project.

El pasado jueves 14 de mayo el profesor Óscar Araque ha presentado en una sesión oral el artículo «AMORES: A Spanish Language Resource for an Extended Set of Moral Foundations» en la conferencia «The Fifteenth Language Resources and Evaluation Conference (LREC 2026)».

 

 

Resumen del trabajo (en Inglés):

This work addresses the need for linguistic resources that enable language models to understand and adapt to subjective and abstract concepts in the domain of moral values within texts. In light of the growing interest in the study of moral values and its limited exploration in Spanish-speaking contexts, this work addresses this gap by developing a novel Spanish-language corpus. Furthermore, the corpus’s development process ensures that the annotations capture a wide range of perspectives, resulting in a resource that reflects the diversity of moral interpretations in real-world contexts. Specifically, there are two main contributions. 1 The creation of the first large-scale Spanish corpus annotated according to Moral Foundations Theory. 2 We introduce an experimental framework that investigates how annotators’ religious orientations could shape moral annotation patterns and propagate to model behavior. To do so, we employ a prompt-based alignment method that improves moral detection regardless of religious alignment for which the model was trained. In this scenario, we explore whether language models can align moral interpretations across divergent belief orientations.

 

Referencia:

Araque, O., Molina, D., Nogales, A. D. A., & Iglesias, C. A. (2026). AMORES: A Spanish Language Resource for an Extended Set of Moral Foundations. In Proceedings of the Fifteenth Language Resources and Evaluation Conference (LREC 2026) (pp. 7056–7068). European Language Resources Association (ELRA). https://doi.org/10.63317/25owsh7mdw6i.

Hoy, 18 de mayo, comienza el proyecto Innovación Digital para la Competitividad de la Cerveza Artesana (GO Cerveza Artesana). El proyecto plantea una transformación integral del subsector de la elaboración de cervezas artesanales, impulsando la digitalización de procesos de elaboración y logística y dotando a las cerveceras de herramientas de marketing digital e IA para mejorar su competitividad a lo largo de toda su cadena de valor.

 

El Grupo de Sistemas Inteligentes participa en el desarrollo de un sistema de inteligencia competitiva para conocer mejor la opinión de los consumidores y las nuevas tendencias en el sector, así como en la creación de asistentes para recomendar cervezas al público. Además, se desarrolla una herramienta inteligente que ayudará  a las empresas de fabricación de cerveza artesana a gestionar sus redes sociales. 


Este sábado 7 de marzo Sergio Muñoz ha presentado el artículo "Evaluation of Diversity in LLM-Based News Discovery Through an Agent-Based System" en el 18th International Conference on Agents and Artificial Intelligence (ICAART). 

Dentro del proyecto AMOR, Fran Cantos ha creado un museo virtual holográfico que permite jugar para mejorar el conocimiento sobre las obras expuestas. El sistema se ha realizado con Grandes Modelos de Lenguaje para generar el juego.