Annotated corpora and tools of the PARSEME Shared Task : Subtask 1, MWE Identification (edition 2.0)

This multilingual resource contains corpora in which multiword expressions (MWEs) of all syntactic categories have been manually annotated. While previous editions covered only verbal MWEs, this version extends the annotation scope to all syntactic MWE categories: verbal, nominal, adjectival, adverbial and functional. This release of the corpora is associated to the PARSEME 2.0 Multilingual Shared Task on Identification and Paraphrasing of Multiword Expressions. It corresponds to the data used in subtask 1, on MWE identification. The data covers 17 languages, of which 7 are new with respect to previous PARSEME editions. The annotation process is based on cross-lingually unified guidelines, phrased as decision diagrams over linguistic tests, and a typology of 18 MWE categories. The corpus contains almost 5 million tokens, over 250,000 sentences and 140,000 MWE annotations. MWEs were annotated according to the universal guidelines. The corpora are provided in the CUPT format, inspired by the CONLL-U format. Morphological and syntactic information, ­­­­including parts of speech, lemmas, morphological features and/or syntactic dependencies, are also provided. Depending on the language, the information comes from treebanks (e.g., Universal Dependencies) or from automatic parsers trained on treebanks (e.g., UDPipe). All corpora are split into training, development and test data, following the splitting strategy adopted for the PARSEME 2.0 Shared Task - subtask 1. This release includes the test data, which was kept secret from participants during the evaluation phase. The annotation guidelines are available online: https://parsemefr.lis-lab.fr/parseme-st-guidelines/2.0 The CUPT format is detailed here: https://gitlab.com/parseme/corpora/-/wikis/CUPT-format Reference ------- When referring to this release, please, cite: * Agata Savary, Manon Scholivet, Carlos Ramisch, Takuya Nakamura, Eric Bilinski, Sara Stymne, Voula Giouli, Stella Markantonatou, Vasile Păiș, Maria Mitrofan, Louis Estève, Bruno Guillaume, Verginica Barbu Mititelu, Jaka Čibej, Roberto A. Díaz Hernández, Victoria Fendel, Polona Gantar, Olha Kanishcheva, Cvetana Krstev, Chaya Liebeskind, Irina Lobzhanidze, Aleksandra Marković, Gunta Nešpore-Bērzkalne, Adriana Pagano, Mehrnoush Shamsfard, Ranka Stanković, Vahide Tajalli, Carole Tiberius, Aakanksha Padhye (2026), _PARSEME 2.0 multilingual corpus of multiword expressions_, in Proceedings of LREC 2026, ELRA, Palma de Mallorca, Spain.