Dataset Preview
Duplicate
The full dataset viewer is not available (click to read why). Only showing a preview of the rows.
The dataset generation failed
Error code:   DatasetGenerationError
Exception:    CastError
Message:      Couldn't cast
doc_id: string
source_url: string
title: struct<es: string, en: string>
  child 0, es: string
  child 1, en: string
journal: struct<name: string, publisher: string, issn_epub: string>
  child 0, name: string
  child 1, publisher: string
  child 2, issn_epub: string
language: string
year: int64
volume: null
issue: string
subjects: list<item: string>
  child 0, item: string
authors: list<item: struct<given: string, family: string, affiliations: list<item: string>, orcid: string>>
  child 0, item: struct<given: string, family: string, affiliations: list<item: string>, orcid: string>
      child 0, given: string
      child 1, family: string
      child 2, affiliations: list<item: string>
          child 0, item: string
      child 3, orcid: string
affiliations: list<item: struct<id: string, name: string, country: string>>
  child 0, item: struct<id: string, name: string, country: string>
      child 0, id: string
      child 1, name: string
      child 2, country: string
license: struct<open_access: bool, url: string, statement: string>
  child 0, open_access: bool
  child 1, url: string
  child 2, statement: string
dates: struct<received: timestamp[s], accepted: timestamp[s], published: string>
  child 0, received: timestamp[s]
  child 1, accepted: timestamp[s]
  child 2, published: string
pages: struct<fpage: string, lpage: string>
  child 0, fpage: string
  child 1, lpage: string
text: string
lang: string
source: string
id: int64
tokens: int64
abstract_en: string
abstract_gl: string
to
{'id': Value('int64'), 'text': Value('string'), 'tokens': Value('int64'), 'abstract_en': Value('string'), 'abstract_gl': Value('string'), 'source': Value('string'), 'lang': Value('string')}
because column names don't match
Traceback:    Traceback (most recent call last):
                File "/usr/local/lib/python3.12/site-packages/datasets/builder.py", line 1779, in _prepare_split_single
                  for key, table in generator:
                                    ^^^^^^^^^
                File "/usr/local/lib/python3.12/site-packages/datasets/packaged_modules/json/json.py", line 295, in _generate_tables
                  self._cast_table(pa_table, json_field_paths=json_field_paths),
                  ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
                File "/usr/local/lib/python3.12/site-packages/datasets/packaged_modules/json/json.py", line 128, in _cast_table
                  pa_table = table_cast(pa_table, self.info.features.arrow_schema)
                             ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
                File "/usr/local/lib/python3.12/site-packages/datasets/table.py", line 2281, in table_cast
                  return cast_table_to_schema(table, schema)
                         ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
                File "/usr/local/lib/python3.12/site-packages/datasets/table.py", line 2227, in cast_table_to_schema
                  raise CastError(
              datasets.table.CastError: Couldn't cast
              doc_id: string
              source_url: string
              title: struct<es: string, en: string>
                child 0, es: string
                child 1, en: string
              journal: struct<name: string, publisher: string, issn_epub: string>
                child 0, name: string
                child 1, publisher: string
                child 2, issn_epub: string
              language: string
              year: int64
              volume: null
              issue: string
              subjects: list<item: string>
                child 0, item: string
              authors: list<item: struct<given: string, family: string, affiliations: list<item: string>, orcid: string>>
                child 0, item: struct<given: string, family: string, affiliations: list<item: string>, orcid: string>
                    child 0, given: string
                    child 1, family: string
                    child 2, affiliations: list<item: string>
                        child 0, item: string
                    child 3, orcid: string
              affiliations: list<item: struct<id: string, name: string, country: string>>
                child 0, item: struct<id: string, name: string, country: string>
                    child 0, id: string
                    child 1, name: string
                    child 2, country: string
              license: struct<open_access: bool, url: string, statement: string>
                child 0, open_access: bool
                child 1, url: string
                child 2, statement: string
              dates: struct<received: timestamp[s], accepted: timestamp[s], published: string>
                child 0, received: timestamp[s]
                child 1, accepted: timestamp[s]
                child 2, published: string
              pages: struct<fpage: string, lpage: string>
                child 0, fpage: string
                child 1, lpage: string
              text: string
              lang: string
              source: string
              id: int64
              tokens: int64
              abstract_en: string
              abstract_gl: string
              to
              {'id': Value('int64'), 'text': Value('string'), 'tokens': Value('int64'), 'abstract_en': Value('string'), 'abstract_gl': Value('string'), 'source': Value('string'), 'lang': Value('string')}
              because column names don't match
              
              The above exception was the direct cause of the following exception:
              
              Traceback (most recent call last):
                File "/src/services/worker/src/worker/job_runners/config/parquet_and_info.py", line 1347, in compute_config_parquet_and_info_response
                  parquet_operations = convert_to_parquet(builder)
                                       ^^^^^^^^^^^^^^^^^^^^^^^^^^^
                File "/src/services/worker/src/worker/job_runners/config/parquet_and_info.py", line 980, in convert_to_parquet
                  builder.download_and_prepare(
                File "/usr/local/lib/python3.12/site-packages/datasets/builder.py", line 882, in download_and_prepare
                  self._download_and_prepare(
                File "/usr/local/lib/python3.12/site-packages/datasets/builder.py", line 943, in _download_and_prepare
                  self._prepare_split(split_generator, **prepare_split_kwargs)
                File "/usr/local/lib/python3.12/site-packages/datasets/builder.py", line 1646, in _prepare_split
                  for job_id, done, content in self._prepare_split_single(
                                               ^^^^^^^^^^^^^^^^^^^^^^^^^^^
                File "/usr/local/lib/python3.12/site-packages/datasets/builder.py", line 1832, in _prepare_split_single
                  raise DatasetGenerationError("An error occurred while generating the dataset") from e
              datasets.exceptions.DatasetGenerationError: An error occurred while generating the dataset

Need help to make the dataset viewer work? Make sure to review how to configure the dataset viewer, and open a discussion for direct support.

id
int64
text
string
tokens
int64
abstract_en
string
abstract_gl
string
source
string
lang
string
10
INNOVACIÓNS REPUBLICANAS NA EDUCACIÓN POPULAR: AS MISIÓNS PEDAGÓXICAS. ANÁLISE DUNHA EXPERIENCIA, CON REFERENCIAS A GALICIA REPUBLICAN INNOVATIONS IN PEOPLE'S EDUCATION: THE MISIONES PEDAGÓGICAS. ANALYSIS OF AN EXPERIENCE, REFERRING TO GALICIA Anxo Serafín Porto Ucha Universidade de Santiago de Compostela angelserafin....
6,995
When the Misiones Pedagógicas arrived in Galicia from 1931 took place a remarkable approach to the Institución Libre de Enseñanza (ILE), born in Madrid in 1876. Despite of being branded as prone to “cultural elitism”, the ILE has a significant dimension focused on popular culture that goes beyond the sphere of mere fee...
Coa chegada das Misións Pedagóxicas a partir de 1931 producíase en Galicia un intenso acercamento á ILE, nacida en Madrid en 1876. A pesar de ser acusada ás veces de “elitismo cultural”, encontramos na ILE unha importante vertente dirixida ao popular que rebasa o ámbito meramente sentimental para converterse en algo má...
1234-121-4788-1-10-20131111.md
es
13
GENERACIÓN DE RESIDUOS DERIVADOS DEL TURISMO RECEPTOR EN GALICIA ANDRÉ CARRASCAL INCERA / MELCHOR FERNÁNDEZ FERNÁNDEZ / XESÚS PEREIRA LÓPEZ Universidad de Santiago de Compostela RECIBIDO: 17 de octubre de 2011 / ACEPTADO: 22 de diciembre de 2011 Resumen: Estimar el impacto del turismo en una sociedad significa consider...
7,395
null
null
1267-61-3541-3-10-20131208.md
es
14
EL TURISMO SOSTENIBLE EN LAS MARCAS TURÍSTICAS DE CATALUÑA TERESA TORRES SOLÉ / MERCÈ SALA RÍOS / MARIONA FARRÉ PERDIGUER Universidad de Lleida RECIBIDO: 20 de octubre de 2011 / ACEPTADO: 22 de diciembre de 2011 Resumen: Las cifras macroeconómicas otorgan al sector turístico una importancia significativa dentro de la e...
9,169
null
null
1269-61-3547-3-10-20131208.md
es
15
"EL MODELO DE FINANCIACIÓN DE LA UNIVERSIDAD GALLEGA 2011-2015: PREVISIÓN DE FONDOS Y BALANCE\nSAR(...TRUNCATED)
8,188
null
null
1270-61-3549-3-10-20131208.md
es
16
"LA CONFIANZA GENERAL EN GALICIA: UNA APROXIMACIÓN A LA EVOLUCIÓN DEL CAPITAL SOCIAL Y DE SUS DETE(...TRUNCATED)
10,165
null
null
1272-61-3555-2-10-20130625.md
es
17
"LOS RESULTADOS DE LA IMPLANTACIÓN DE LA NORMA EN 9100 EN EL SECTOR AEROESPACIAL ESPAÑOL\nCAMILO P(...TRUNCATED)
9,682
null
null
1273-61-3558-3-10-20131208.md
es
18
"GRADO DE CUMPLIMIENTO POR LOS AUDITORES DE LAS NORMAS DE AUDITORÍA SOBRE INFORMES. UN ESTUDIO EMP(...TRUNCATED)
9,501
null
null
1275-61-3564-3-10-20131208.md
es
19
"AUDITORÍA, COMITÉS DE AUDITORÍA Y NEUTRALIDAD EN LA INFORMACIÓN NARRATIVA\nMARÍA ROSARIO BABÍ(...TRUNCATED)
10,622
null
null
1276-61-3567-3-10-20131208.md
es
20
"POLÍTICAS EUROPEAS SOBRE AUDITORÍA EN TIEMPOS DE CRISIS\nMARÍA ANTONIA GARCÍA BENAU* / ANA ZORI(...TRUNCATED)
9,089
null
null
1277-61-3570-2-10-20130625.md
es
21
"LOS REQUISITOS EDUCATIVOS PARA EL ACCESO Y LA FORMACIÓN CONTINUADA DE LOS AUDITORES\nJosé A. Gonz(...TRUNCATED)
13,986
null
null
1281-61-3580-3-10-20131208.md
es
End of preview.

Corpus de dominio científico y académico

Descripción general

El corpus de dominio científico reúne textos académicos y enciclopédicos representativos del registro científico, procedentes de fuentes institucionales y abiertas. Incluye publicaciones universitarias, artículos científicos y contenidos enciclopédicos, con el objetivo de proporcionar un recurso útil para tareas de procesamiento del lenguaje natural. El corpus se encuentra además en proceso de ampliación mediante la incorporación de nuevas publicaciones científicas, por lo que se prevén futuras actualizaciones del recurso.

Composición del corpus

El dominio científico está formado por los siguientes conjuntos:

Publicaciones del Servicio de Publicaciones de la Universidad de Santiago de Compostela (USC)

Este conjunto incluye artículos y publicaciones académicas editadas por la USC, obtenidas a partir de dos tipos de fuentes:

  • Documentos extraídos de XML: cuando los artículos estaban disponibles en formato XML estructurado, se extrajo el texto completo junto con los metadatos editoriales disponibles, como el título, los autores, el idioma, el año de publicación o identificadores como el DOI. Los documentos resultantes se almacenan en formato JSONL.
  • Documentos extraídos de PDF mediante OCR: en los casos en los que no existía una versión XML accesible, los textos se obtuvieron a partir de archivos PDF utilizando PaddleOCR. Estos documentos contienen principalmente el texto del artículo tras OCR, con metadatos mínimos, y fueron sometidos a un proceso posterior de limpieza.

Estos textos están incluidos en CorpusNÓS, pero han pasado por un proceso de extracción y limpieza diferente.

Wikipedia (ámbito científico)

El corpus incluye artículos de Wikipedia pertenecientes al ámbito científico, tanto en castellano como en gallego, recopilados mediante volcados temáticos específicos por idioma.

Para cada artículo se extrajo el texto enciclopédico limpio, junto con información básica como el título, la URL y la categoría temática. Este conjunto aporta diversidad temática y un registro más divulgativo dentro del dominio científico, complementando los textos académicos.

Formato y organización

Todos los documentos del dominio científico se almacenan en formato JSONL, con un documento por línea. La organización del corpus refleja la fuente de procedencia, el idioma y el formato original del texto (XML o PDF).

La estructura del corpus es la siguiente:

corpus_dominio_cientifico/
├── es/
│   ├── Wikipedia/
│   │   └── wikipedia_ciencia_es_13112025_dump.jsonl
│   └── Servizo_Publicacions_USC/
│       ├── source_pdf/
│       │   └── ServizoPublicacionsUSC_es.jsonl
│       └── source_xml/
│           ├── 2020.jsonl
│           ├── 2021.jsonl
│           └── 2022.jsonl
└── gl/
    ├── Wikipedia/
    │   └── wikipedia_ciencia_gl_11112025_dump.jsonl
    └── Servizo_Publicacions_USC/
        ├── source_pdf/
        │   └── ServizoPublicacionsUSC_gl.jsonl
        └── source_xml/
            ├── 2021.jsonl
            ├── 2022.jsonl
            └── 2023.jsonl

Procesamiento y limitaciones

El corpus ha sido generado mediante procesos automáticos de extracción y limpieza básica del texto. No se ha realizado corrección lingüística manual ni anotación semántica adicional.

Los textos procedentes de PDF pueden contener errores residuales derivados del proceso de OCR. Asimismo, la disponibilidad y riqueza de metadatos varía en función de la fuente y del formato original.

Ejemplos

Wikipedia

{
  "id": 1,
  "title": "%Alcons",
  "text": "% Alcons es un acrónimo de Percentage Articulation Loss of Consonants...",
  "related_articles": [],
  "num_tokens": 384,
  "url": "https://es.wikipedia.org/wiki/%25Alcons"
}

Servicio de Publicaciones de la USC (XML)

{
  "doc_id": "10.15304/ges.2.6469",
  "source_url": "https://doi.org/10.15304/ges.2.6469",
  "title": {
    "es": "LA PROTECCIÓN PENAL FRENTE A LA MINERÍA ILEGAL"
  },
  "journal": {
    "name": "Gladius et Scientia. Revista de Seguridad del CESEG",
    "publisher": "Universidade de Santiago de Compostela",
    "issn_epub": "2695-6993"
  },
  "language": "es",
  "year": 2020,
  "authors": [
    {
      "given": "Antonio",
      "family": "Roma Valdés"
    }
  ],
  "text": "1. INTRODUCCIÓN\n\nLa minería ilegal es un fenómeno..."
}

Servicio de Publicaciones de la USC (PDF)

  "id": 10,
  "text": "INNOVACIÓNS REPUBLICANAS NA EDUCACIÓN POPULAR...",
  "tokens": 6995,
  "abstract_es": null,
  "abstract_gl": "Coa chegada das Misións Pedagóxicas a partir de 1931...",
  "abstract_en": "When the Misiones Pedagógicas arrived in Galicia...",
  "source": "1234-121-4788-1-10-20131111.md",
  "lang": "es"
}

Información adicional

Fuente Idioma Nº de documentos Nº de tokens
Servizo de Publicacións da USC Español (es) 186 1,307,411
Servizo de Publicacións da USC Gallego (gl) 499 3,005,739
Wikipedia (ciencia) Español (es) 34,151 33,957,522
Wikipedia (ciencia) Gallego (gl) 35,964 20,448,504
Total 70,800 58,719,176

Usos previstos

Este conjunto de datos puede utilizarse para:

  • modelado de lenguaje específico del dominio científico
  • preentrenamiento continuo sobre texto académico y enciclopédico
  • análisis de corpus del registro científico en gallego y español
  • experimentos bilingües o translingües con textos científicos
  • investigación en PLN para lenguas de bajos recursos y lenguajes especializados

Licencia y condiciones de uso

Publicaciones del Servicio de Publicaciones de la Universidad de Santiago de Compostela (USC)

Los textos procedentes del Servicio de Publicaciones de la Universidad de Santiago de Compostela han sido incluidos en este corpus en el marco de los acuerdos institucionales vigentes.

Los documentos originales en formato XML han sido reformateados para su almacenamiento en JSONL, preservando el contenido textual y los metadatos editoriales disponibles, sin alteración del contenido semántico. Los textos extraídos de PDF han sido procesados mediante OCR y limpieza básica.

Wikipedia

Los textos enciclopédicos proceden de Wikipedia en castellano y gallego y han sido extraídos de conformidad con la licencia Creative Commons Attribution–ShareAlike 4.0 International (CC BY-SA 4.0) bajo la que se publica Wikipedia.

Los artículos han sido transformados mediante procesos automáticos de extracción, limpieza y reestructuración en formato JSONL, sin modificar su contenido textual. De acuerdo con los términos de la licencia original, se mantiene la atribución a Wikipedia como fuente y se preserva la compatibilidad con su licencia.

Licencia del corpus derivado

La estructura del corpus, el formato de los datos, la organización de los archivos y los procesos de extracción, limpieza y normalización aplicados se distribuyen bajo la licencia:

Creative Commons Attribution 4.0 International (CC BY 4.0)

Referencias

@misc{cui2025paddleocr30technicalreport,
      title={PaddleOCR 3.0 Technical Report}, 
      author={Cheng Cui and Ting Sun and Manhui Lin and Tingquan Gao and Yubo Zhang and Jiaxuan Liu and Xueqing Wang and Zelun Zhang and Changda Zhou and Hongen Liu and Yue Zhang and Wenyu Lv and Kui Huang and Yichao Zhang and Jing Zhang and Jun Zhang and Yi Liu and Dianhai Yu and Yanjun Ma},
      year={2025},
      eprint={2507.05595},
      archivePrefix={arXiv},
      primaryClass={cs.CV},
      url={https://arxiv.org/abs/2507.05595}
}

@misc{cui2025paddleocrvlboostingmultilingualdocument,
      title={PaddleOCR-VL: Boosting Multilingual Document Parsing via a 0.9B Ultra-Compact Vision-Language Model}, 
      author={Cheng Cui and Ting Sun and Suyin Liang and Tingquan Gao and Zelun Zhang and Jiaxuan Liu and Xueqing Wang and Changda Zhou and Hongen Liu and Manhui Lin and Yue Zhang and Yubo Zhang and Handong Zheng and Jing Zhang and Jun Zhang and Yi Liu and Dianhai Yu and Yanjun Ma},
      year={2025},
      eprint={2510.14528},
      archivePrefix={arXiv},
      primaryClass={cs.CV},
      url={https://arxiv.org/abs/2510.14528}
}

@inproceedings{de-dios-flores-etal-2024-corpusnos,
    title = "{C}orpus{N{\'O}S}: A massive {G}alician corpus for training large language models",
    author = "de-Dios-Flores, Iria  and
      Su{\'a}rez, Silvia Paniagua  and
      P{\'e}rez, Cristina Carbajal  and
      Outeiri{\~n}o, Daniel Bardanca  and
      Garcia, Marcos  and
      Gamallo, Pablo",
    editor = "Gamallo, Pablo  and
      Claro, Daniela  and
      Teixeira, Ant{\'o}nio  and
      Real, Livy  and
      Garcia, Marcos  and
      Oliveira, Hugo Gon{\c{c}}alo  and
      Amaro, Raquel",
    booktitle = "Proceedings of the 16th International Conference on Computational Processing of Portuguese - Vol. 1",
    month = mar,
    year = "2024",
    address = "Santiago de Compostela, Galicia/Spain",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/2024.propor-1.66/",
    pages = "593--599"
}

Financiación y agradecimientos

Esta publicación del proyecto Desarrollo de Modelos ALIA está financiada por el Ministerio para la Transformación Digital y de la Función Pública y por el Plan de Recuperación, Transformación y Resiliencia – Financiado por la Unión Europea – NextGenerationEU.

This work is funded by the Ministerio para la Transformación Digital y de la Función Pública - Funded by EU – NextGenerationEU within the framework of the project Desarrollo de Modelos ALIA.

Downloads last month
89

Models trained or fine-tuned on proxectonos/corpus_dominio_cientifico

Collection including proxectonos/corpus_dominio_cientifico

Papers for proxectonos/corpus_dominio_cientifico