DIGITAL.CSIC. Repositorio Institucional del CSIC
Research data
Set de datos (Dataset)
CLARA-MeD corpus
A collection of 24.298 pairs of professional and simplified texts (>96 million tokens): 1) Drug leaflets and summaries of product characteristics (10 211 pairs of texts, >82M words); 2) Cancer-related information summaries (201 pairs of texts, >3M tokens); and 2) Clinical trials announcements (5748 pairs of texts, 451 690 tokens).
The dataset also contains a parallel corpus with a subset of 3800 sentence pairs of professional and laymen variants (149 862 tokens). This is a benchmark for medical text simplification. The latest download of files was in February 2022., The data can be reproduced (partially or totally), as long as the source is provided. The specific details of each data source are as follows: - For the data from CIMA/AEMPS: The copyright of the contents belongs to Centro de Información de Medicamentos (CIMA, 'Medicine Online Information Center'), from the Agencia Española de Medicamentos y Productos Sanitarios (AEMPS). The latest access to the data was February 2022 from https://cima.aemps.es. Note that CIMA updates its contents regularly, and the data in the CLARA-MeD corpus might not be updated according to the version of the online contents. The data can be reproduced (partially or totally), as long as the source is provided. For more information, please, check: https://www.aemps.gob.es/avisoLegal/ - For the data from EudraCT: We attribute the source as "EU Clinical Trials Register" in the publication or distribution of these data, and clearly state that these data were accessed as of January 2022 from https://www.clinicaltrialsregister.eu The European Medicines Agency, according to current European Union and international legislation, is the owner of copyright and other intellectual property rights for documents and other content published on its website. Information and documents made available on the Agency's webpages are public and may be reproduced and/or distributed, totally or in part, irrespective of the means and/or the formats used, for non-commercial and commercial purposes, provided that the Agency is always acknowledged as the source of the material. Such acknowledgement must be included in each copy of the material. For further information, please, see: https://www.ema.europa.eu/en/about-us/legal-notice - For the data from the NCI: Unless otherwise indicated, all text within NCI products is free of copyright and may be reused without permission. Credit the National Cancer Institute as the source. For further information, please, see: https://www.cancer.gov/policies/copyright-reuse, A collection of 24 298 pairs of professional and simplified texts (>96 million tokens) for automatic medical text simplification in Spanish.
A parallel corpus with a subset of 3800 sentence pairs of professional and laymen variants (149 862 tokens) is released as a benchmark for medical text simplification.
This dataset was collected in the CLARA-MeD project, with the goal of simplifying medical texts in the Spanish language and reducing the language barrier to patient's informed decision making. In particular, the project aims at developing linguistic resources for automatic medical term simplification in Spanish; and conducting experiments in automatic text simplification., This dataset was collected in the CLARA-MeD project (PID2020- 116001RA-C33), with funding from the Spanish government by MCIN/AEI/10.13039/501100011033/, in project call: “Proyectos I+D+i Retos Investigación”., Folders: 1) Comparable corpus: the data of each source can be found in the corresponding folder. Each folder contains two other subfolders: - source: professional, specialized texts (".src" file extension) - target: simplified texts (".trg" file extension). 2) Aligned sentences: these can be found file "aligned.tsv" (in folder "aligned"). The folder structure is as follows: - aligned/ - aligned.tsv - cima/ - source/ - target/ - eudract/ - source/ - target/ - nci/ - source/ - target/., Peer reviewed