Too Young to NER: Improving Entity Recognition on Dutch Historical Documents

Vera Provatorova; Marieke van Erp; Evangelos Kanoulas

Too Young to NER: Improving Entity Recognition on Dutch Historical Documents

Vera Provatorova, Marieke van Erp, Evangelos Kanoulas

DHLab

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

Abstract

Named entity recognition (NER) on historical texts is beneficial for the field of digital humanities, as it allows to easily search for the names of people, places and other entities in digitised archives. While the task of historical NER in different languages has been gaining popularity in recent years, Dutch historical NER remains an underexplored topic. Using a recently released historical dataset from the Dutch Language Institute, we train three BERT-based models and analyse the errors to identify main challenges. All three models outperform a contemporary multilingual baseline by a large margin on historical test data.

Original language	English
Pages	30-35
Number of pages	6
Publication status	Published - 01 May 2024

Access to Document

https://aclanthology.org/2024.lt4hala-1.4.pdf

Cite this

@conference{7ea1014401244c779b420a6e4aa8cf4c,

title = "Too Young to NER: Improving Entity Recognition on Dutch Historical Documents",

abstract = "Named entity recognition (NER) on historical texts is beneficial for the field of digital humanities, as it allows to easily search for the names of people, places and other entities in digitised archives. While the task of historical NER in different languages has been gaining popularity in recent years, Dutch historical NER remains an underexplored topic. Using a recently released historical dataset from the Dutch Language Institute, we train three BERT-based models and analyse the errors to identify main challenges. All three models outperform a contemporary multilingual baseline by a large margin on historical test data.",

author = "Vera Provatorova and {van Erp}, Marieke and Evangelos Kanoulas",

year = "2024",

month = may,

day = "1",

language = "English",

pages = "30--35",

}

TY - CONF

T1 - Too Young to NER: Improving Entity Recognition on Dutch Historical Documents

AU - Provatorova, Vera

AU - van Erp, Marieke

AU - Kanoulas, Evangelos

PY - 2024/5/1

Y1 - 2024/5/1

N2 - Named entity recognition (NER) on historical texts is beneficial for the field of digital humanities, as it allows to easily search for the names of people, places and other entities in digitised archives. While the task of historical NER in different languages has been gaining popularity in recent years, Dutch historical NER remains an underexplored topic. Using a recently released historical dataset from the Dutch Language Institute, we train three BERT-based models and analyse the errors to identify main challenges. All three models outperform a contemporary multilingual baseline by a large margin on historical test data.

AB - Named entity recognition (NER) on historical texts is beneficial for the field of digital humanities, as it allows to easily search for the names of people, places and other entities in digitised archives. While the task of historical NER in different languages has been gaining popularity in recent years, Dutch historical NER remains an underexplored topic. Using a recently released historical dataset from the Dutch Language Institute, we train three BERT-based models and analyse the errors to identify main challenges. All three models outperform a contemporary multilingual baseline by a large margin on historical test data.

M3 - Paper

SP - 30

EP - 35

ER -