Repurposing Holocaust-Related Digital Scholarly Editions to Develop Multilingual Domain-Specific Named Entity Recognition Tools

Maria Dermentzi; Hugo Scheithauer; Isuri Anuradha; Martin Wynne; Francesca Frontini; Alistair Plum

Repurposing Holocaust-Related Digital Scholarly Editions to Develop Multilingual Domain-Specific Named Entity Recognition Tools

Maria Dermentzi, Hugo Scheithauer, Isuri Anuradha (Editor), Martin Wynne (Editor), Francesca Frontini (Editor), Alistair Plum (Editor)

Digital Humanities

Research output: Contribution to conference types › Paper › peer-review

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%. We argue that this score is sufficiently high to consider the next steps towards deploying this model.

Original language	English
Pages	18-28
Number of pages	11
Publication status	Published - May 2024

Access to Document

https://aclanthology.org/2024.htres-1.3

Cite this

@conference{bcee927d7197494789855f2764253b0b,

title = "Repurposing Holocaust-Related Digital Scholarly Editions to Develop Multilingual Domain-Specific Named Entity Recognition Tools",

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%. We argue that this score is sufficiently high to consider the next steps towards deploying this model.",

author = "Maria Dermentzi and Hugo Scheithauer and Isuri Anuradha and Martin Wynne and Francesca Frontini and Alistair Plum",

note = "Publisher Copyright: {\textcopyright} 2024 ELRA Language Resource Association.",

year = "2024",

month = may,

language = "English",

pages = "18--28",

}

TY - CONF

T1 - Repurposing Holocaust-Related Digital Scholarly Editions to Develop Multilingual Domain-Specific Named Entity Recognition Tools

AU - Dermentzi, Maria

AU - Scheithauer, Hugo

A2 - Anuradha, Isuri

A2 - Wynne, Martin

A2 - Frontini, Francesca

A2 - Plum, Alistair

PY - 2024/5

Y1 - 2024/5

N2 - 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%. We argue that this score is sufficiently high to consider the next steps towards deploying this model.

AB - 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%. We argue that this score is sufficiently high to consider the next steps towards deploying this model.

UR - http://www.scopus.com/inward/record.url?scp=85195148766&partnerID=8YFLogxK

M3 - Paper

SP - 18

EP - 28

ER -

Repurposing Holocaust-Related Digital Scholarly Editions to Develop Multilingual Domain-Specific Named Entity Recognition Tools

Abstract

Access to Document

Other files and links

Fingerprint

Cite this