IC-UNED-RC-ES: Spanish Reading Comprehension Dataset developed by Instituto Cervantes and UNED

IC-UNED-RC-ES is a dataset Reading Comprehension exercises developed by Instituto Cervantes to assess their students language proficiency in Spanish. UNED has produced its electronic version to enable the evaluation of Artificial Intelligence systems. The dataset contains different types of exercises at different levels of difficulty. It also contains images in some of the questions and answers. This dataset has been used in the Shared Task evaluation PROFE 2025 at IberLEF: https://nlp.uned.es/question-answering/profe2025 The dataset is provided without the keys so the large language models can't be contaminated. If you are interested on testing your system, please visite PROFE 2025 site or write [e-mail]. This collaboration has been possible in the framework of PROYECTO DE CREACIÓN DE UN CORPUS DIGITALIZADO DE TAREAS DE COMPRENSIÓN DE LECTURA PARA LA EVALUACIÓN DE LA COMPRENSIÓN EN SISTEMAS DE INTELIGENCIA ARTIFICIAL. UNED has been partially funded by DeepInfo Project (AEI PID2021-127777OB-C22)