Datasets:

ehri-ner
/

ehri-ner-all

Error code:   DatasetGenerationError
Exception:    ArrowNotImplementedError
Message:      Cannot write struct type '_format_kwargs' with no child field to Parquet. Consider adding a dummy child field.
Traceback:    Traceback (most recent call last):
                File "/src/services/worker/.venv/lib/python3.9/site-packages/datasets/builder.py", line 2011, in _prepare_split_single
                  writer.write_table(table)
                File "/src/services/worker/.venv/lib/python3.9/site-packages/datasets/arrow_writer.py", line 583, in write_table
                  self._build_writer(inferred_schema=pa_table.schema)
                File "/src/services/worker/.venv/lib/python3.9/site-packages/datasets/arrow_writer.py", line 404, in _build_writer
                  self.pa_writer = self._WRITER_CLASS(self.stream, schema)
                File "/src/services/worker/.venv/lib/python3.9/site-packages/pyarrow/parquet/core.py", line 1010, in __init__
                  self.writer = _parquet.ParquetWriter(
                File "pyarrow/_parquet.pyx", line 2157, in pyarrow._parquet.ParquetWriter.__cinit__
                File "pyarrow/error.pxi", line 154, in pyarrow.lib.pyarrow_internal_check_status
                File "pyarrow/error.pxi", line 91, in pyarrow.lib.check_status
              pyarrow.lib.ArrowNotImplementedError: Cannot write struct type '_format_kwargs' with no child field to Parquet. Consider adding a dummy child field.
              
              During handling of the above exception, another exception occurred:
              
              Traceback (most recent call last):
                File "/src/services/worker/.venv/lib/python3.9/site-packages/datasets/builder.py", line 2027, in _prepare_split_single
                  num_examples, num_bytes = writer.finalize()
                File "/src/services/worker/.venv/lib/python3.9/site-packages/datasets/arrow_writer.py", line 602, in finalize
                  self._build_writer(self.schema)
                File "/src/services/worker/.venv/lib/python3.9/site-packages/datasets/arrow_writer.py", line 404, in _build_writer
                  self.pa_writer = self._WRITER_CLASS(self.stream, schema)
                File "/src/services/worker/.venv/lib/python3.9/site-packages/pyarrow/parquet/core.py", line 1010, in __init__
                  self.writer = _parquet.ParquetWriter(
                File "pyarrow/_parquet.pyx", line 2157, in pyarrow._parquet.ParquetWriter.__cinit__
                File "pyarrow/error.pxi", line 154, in pyarrow.lib.pyarrow_internal_check_status
                File "pyarrow/error.pxi", line 91, in pyarrow.lib.check_status
              pyarrow.lib.ArrowNotImplementedError: Cannot write struct type '_format_kwargs' with no child field to Parquet. Consider adding a dummy child field.
              
              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 1529, 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 1154, in convert_to_parquet
                  builder.download_and_prepare(
                File "/src/services/worker/.venv/lib/python3.9/site-packages/datasets/builder.py", line 1027, in download_and_prepare
                  self._download_and_prepare(
                File "/src/services/worker/.venv/lib/python3.9/site-packages/datasets/builder.py", line 1122, in _download_and_prepare
                  self._prepare_split(split_generator, **prepare_split_kwargs)
                File "/src/services/worker/.venv/lib/python3.9/site-packages/datasets/builder.py", line 1882, in _prepare_split
                  for job_id, done, content in self._prepare_split_single(
                File "/src/services/worker/.venv/lib/python3.9/site-packages/datasets/builder.py", line 2038, 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.

_data_files list	_fingerprint string	_format_columns sequence	_format_kwargs dict	_format_type null	_output_all_columns bool	_split null
[ { "filename": "data-00000-of-00001.arrow" } ]	410c67fce89a2c76	[ "id", "tokens", "ner_tags" ]	{}	null	false	null

Dataset Card for ehri-ner/ehri-ner-all

The European Holocaust Research Infrastructure (EHRI) aims to support Holocaust research by making information about dispersed Holocaust material accessible and interconnected through its services. Creating a tool capable of detecting named entities in texts such as Holocaust testimonies or archival descriptions would make it easier to link more material with relevant identifiers in domain-specific controlled vocabularies, semantically enriching it, and making it more discoverable. The EHRI-NER dataset is a multilingual dataset (Czech, German, English, French, Hungarian, Dutch, Polish, Slovak, Yiddish) suitable for training domain-specific Named Entity Recognition (NER) models for Holocaust-related texts.

We have converted all available Extensible Markup Language (XML) files from the EHRI digital scholarly editions (i.e., EHRI Online Editions) into a corpus in a format suitable for training NER models.

Dataset Details

The EHRI-NER dataset includes a total of 505758 tokens, with 5351 person entities, 9399 location entities, 1867 organization entities, 2237 date entities, 528 ghetto entities, and 1229 camp entities.

Dataset Description

Since 2018, the EHRI Consortium has supported the development and publication of six Holocaust-related digital scholarly editions (see here). Each edition enables digital access to facsimiles and transcripts of thematically related documents held by different EHRI partner institutions through a single web interface and unlocks new ways of presenting and browsing through historical sources using digital tools. Publishing a digital edition is a resource-intensive process. Notwithstanding the extensive archival research needed for selecting the documents, additional steps include transcribing and translating them and, most importantly, annotating words and phrases found within these texts and creating links with entities in controlled vocabularies provided by EHRI and third parties. Currently, this annotation is done manually by or under the supervision of subject matter experts, ensuring a high quality of annotations. We repurposed these resources to convert them into a dataset suitable for training NER models, which we consider as a gold standard.

Each EHRI Online Edition consists of digitized documents originating from various archives that are selected, edited, and annotated by EHRI researchers using the Text Encoding Initiative (TEI) P5 standard, an XML schema, which supports their online publication. Editions enhance the edited documents by contextualizing the information contained within them and linking them to EHRI vocabularies and descriptions, and by visualizing georeferenced entities through interactive maps. Thanks to their encoding in TEI, they are fully searchable and can be filtered using facets such as spatial locations, topics, persons, organizations, and institutions. All documents within an edition have a transcript, either in their original language, a translation, or both, and have access to their facsimile. EHRI Editions are published without a regular schedule and it is possible to update them with new material or improve the already published documents.

The resulting EHRI-NER dataset includes nine languages: Czech (cs), German (de), English (en), French (fr), Hungarian (hu), Dutch (nl), Polish (pl), Slovak (sk), and Yiddish (yi).

Annotation format

Each word has been put on a separate line and there is an empty line after each sentence.

The annotations follow the conll2003 format (IOB).

Entity Classes

PER, LOC, ORG, DATE, CAMP, GHETTO

Curated by: EHRI
Funded by: European Commission call H2020-INFRAIA-2018–2020. Grant agreement ID 871111. DOI 10.3030/871111.
Shared by: Dermentzi, M. & Scheithauer, H.
Language(s) (NLP): cs, de, en, fr, hu, nl, pl, sk, yi
License: EUPL-1.2

Dataset Sources

Repository: https://github.com/EHRI/EHRI-NER
Paper: https://hal.science/hal-04547222

Uses

EHRI-NER is a multilingual dataset (Czech, German, English, French, Hungarian, Dutch, Polish, Slovak, Yiddish) for Named Entity Recognition (NER) in Holocaust-related texts. EHRI-NER is built by aggregating all the annotated documents in the EHRI Online Editions and converting them into a format suitable for training domain-specific NER models.

Source Data

This dataset is derived from the EHRI Online Editions, a series of six Holocaust-related digital scholarly editions (more info here).

Who are the source data producers?

This dataset was made possible thanks to the previous work of the editors and contributors of the EHRI Online Editions, including the annotators, the people who produced digital facsimiles of the original archival material, and those who created the transcripts and translations.

Limitations

This dataset stems from a series of manually annotated digital scholarly editions, the EHRI Online Editions. The original purpose of these editions was not to provide a dataset for training NER models, although we argue that they nevertheless constitute a high-quality resource that is suitable to be used in this way. However, users should still be mindful that our dataset repurposes a resource that was not built for purpose.

This dataset was put together as part of an EHRI-specific research project and may not be suitable for the purposes of other users/organizations.

Recommendations

For more information, we encourage potential users to read the paper accompanying this dataset: Dermentzi, M., & Scheithauer, H. (2024, May). Repurposing Holocaust-Related Digital Scholarly Editions to Develop Multilingual Domain-Specific Named Entity Recognition Tools. LREC-COLING 2024 - Joint International Conference on Computational Linguistics, Language Resources and Evaluation. HTRes@LREC-COLING 2024, Torino, Italy. https://hal.science/hal-04547222

Citation

BibTeX: @inproceedings{dermentzi_repurposing_2024, address = {Torino, Italy}, title = {Repurposing {Holocaust}-{Related} {Digital} {Scholarly} {Editions} to {Develop} {Multilingual} {Domain}-{Specific} {Named} {Entity} {Recognition} {Tools}}, url = {https://hal.science/hal-04547222}, abstract = {The European Holocaust Research Infrastructure (EHRI) aims to support Holocaust research by making information about dispersed Holocaust material accessible and interconnected through its services. Creating a tool capable of detecting named entities in texts such as Holocaust testimonies or archival descriptions would make it easier to link more material with relevant identifiers in domain-specific controlled vocabularies, semantically enriching it, and making it more discoverable. With this paper, we release EHRI-NER, a multilingual dataset (Czech, German, English, French, Hungarian, Dutch, Polish, Slovak, Yiddish) for Named Entity Recognition (NER) in Holocaust-related texts. EHRI-NER is built by aggregating all the annotated documents in the EHRI Online Editions and converting them to a format suitable for training NER models. We leverage this dataset to fine-tune the multilingual Transformer-based language model XLM-RoBERTa (XLM-R) to determine whether a single model can be trained to recognize entities across different document types and languages. The results of our experiments show that despite our relatively small dataset, in a multilingual experiment setup, the overall F1 score achieved by XLM-R fine-tuned on multilingual annotations is 81.5{\textbackslash}%. We argue that this score is sufficiently high to consider the next steps towards deploying this model.}, urldate = {2024-04-29}, booktitle = {{LREC}-{COLING} 2024 - {Joint} {International} {Conference} on {Computational} {Linguistics}, {Language} {Resources} and {Evaluation}}, publisher = {ELRA Language Resources Association (ELRA); International Committee on Computational Linguistics (ICCL)}, author = {Dermentzi, Maria and Scheithauer, Hugo}, month = may, year = {2024}, keywords = {Digital Editions, Holocaust Testimonies, Multilingual, Named Entity Recognition, Transfer Learning, Transformers}, }

APA: Dermentzi, M., & Scheithauer, H. (2024, May). Repurposing Holocaust-Related Digital Scholarly Editions to Develop Multilingual Domain-Specific Named Entity Recognition Tools. LREC-COLING 2024 - Joint International Conference on Computational Linguistics, Language Resources and Evaluation. HTRes@LREC-COLING 2024, Torino, Italy. https://hal.science/hal-04547222

Downloads last month: 80

Models trained or fine-tuned on ehri-ner/ehri-ner-all

ehri-ner/xlm-roberta-large-ehri-ner-all

Token Classification • Updated Apr 29 • 113 • 2