Silvia Casonal receives Marie Curie 2025 - Next september TALN will welcome Silvia Casola, recently awarded a Marie Curie 2025 award. For her project titled: Linking Generation and Evaluation for Reliable NLG Assessment - GenEval under the supervision of Horacio Saggion, Professor of Computer Science and Artificial Intelligence and Director of the Natural Language Processing Research Group (TALN) of the Department of Engineering
Amount awarded: €194,074.56

Abstract of the project

GenEval is a project to evaluate the reliability of AI tools for assessing text correction. Currently, generative AI tools do not have the same skills for generating texts as for assessing them and there are many unknowns about the internal mechanisms that link both functions. This research aims to generate new datasets and metrics to determine the degree of reliability of an AI as an evaluator and how it could be combined with human supervision. For Silvia Casola, "the tools that will be developed in GenEval will play an essential role in fostering greater trust in AI systems and ensuring more reliable and transparent performance assessments, which could benefit a wide range of professionals, given the growing role of language technologies in critical areas such as health, law or education."