Overview of MER-TRANS at IberLEF 2026: First Shared Task on Multilingual Easy-to-Read Translation

Resumen

This paper presents MER-TRANS, a shared task on multilingual easy-to-read translation organised within the context of IberLEF, the evaluation campaign for Natural Language Processing (NLP) systems in Spanish and other Iberian languages. The objective of MER-TRANS is to challenge the NLP community to develop solutions that produce easy-to-read versions of texts and sentences in Arabic, Catalan, Italian, and Spanish. The task features a multilingual corpus of original documents and their easy-to-read adaptations in Catalan, Italian, and Spanish, as well as a sentence-aligned dataset of original sentences and their simplifications in Arabic. MER-TRANS attracted seven teams which submitted a total of 42 systems (including the organisers' baselines), all of which were evaluated using automatic metrics.