Although pre-trained named entity recognition (NER) models are highly accurate on modern corpora, they underperform on historical texts due to differences in language OCR errors. In this work, we develop a new NER corpus of 3.6M sentences from late medieval charters written mainly in Czech, Latin, and German.
We show that we can start with a list of known historical figures and locations and an unannotated corpus of historical texts, and use information retrieval techniques to automatically bootstrap a NER-annotated corpus. Using our corpus, we train a NER model that achieves entity-level Precision of 72.81-93.98% with 58.14-81.77% Recall on a manually-annotated test dataset. Furthermore, we show that using a weighted loss function helps to combat class imbalance in token classification tasks. To make it easy for others to reproduce and build upon our work, we publicly release our corpus, models, and experimental code.
Návaznosti
MUNI/A/1339/2022, interní kód MU
Název: Rozvoj technik pro zpracování dat pro podporu vyhledávání, analýz a vizualizací rozsáhlých datových souborů s využitím umělé inteligence
Investor: Masarykova univerzita, Rozvoj technik pro zpracování dat pro podporu vyhledávání, analýz a vizualizací rozsáhlých datových souborů s využitím umělé inteligence
NOVOTNÝ, Vít, Kristýna LUGER, Michal ŠTEFÁNIK, Tereza VRABCOVÁ a Aleš HORÁK. People and Places of Historical Europe: Bootstrapping Annotation Pipeline and a New Corpus of Named Entities in Late Medieval Texts. Online. In Anna Rogers, Jordan Boyd-Graber, Naoaki Okazaki. Findings of the Association for Computational Linguistics: ACL 2023. Toronto, Canada: Association for Computational Linguistics, 2023, s. 14104-14113. ISBN 978-1-959429-62-3.
@inproceedings{2289944, author = {Novotný, Vít and Luger, Kristýna and Štefánik, Michal and Vrabcová, Tereza and Horák, Aleš}, address = {Toronto, Canada}, booktitle = {Findings of the Association for Computational Linguistics: ACL 2023}, editor = {Anna Rogers, Jordan Boyd-Graber, Naoaki Okazaki}, keywords = {natural language processing; nlp; historical documents; optical character recognition; ocr; named entity recognition; ner; czech; german; latin}, howpublished = {elektronická verze "online"}, language = {eng}, location = {Toronto, Canada}, isbn = {978-1-959429-62-3}, pages = {14104-14113}, publisher = {Association for Computational Linguistics}, title = {People and Places of Historical Europe: Bootstrapping Annotation Pipeline and a New Corpus of Named Entities in Late Medieval Texts}, url = {https://aclanthology.org/2023.findings-acl.887}, year = {2023} }
TY - JOUR ID - 2289944 AU - Novotný, Vít - Luger, Kristýna - Štefánik, Michal - Vrabcová, Tereza - Horák, Aleš PY - 2023 TI - People and Places of Historical Europe: Bootstrapping Annotation Pipeline and a New Corpus of Named Entities in Late Medieval Texts PB - Association for Computational Linguistics CY - Toronto, Canada SN - 9781959429623 KW - natural language processing KW - nlp KW - historical documents KW - optical character recognition KW - ocr KW - named entity recognition KW - ner KW - czech KW - german KW - latin UR - https://aclanthology.org/2023.findings-acl.887 N2 -
Although pre-trained named entity recognition (NER) models are highly accurate on modern corpora, they underperform on historical texts due to differences in language OCR errors. In this work, we develop a new NER corpus of 3.6M sentences from late medieval charters written mainly in Czech, Latin, and German.
We show that we can start with a list of known historical figures and locations and an unannotated corpus of historical texts, and use information retrieval techniques to automatically bootstrap a NER-annotated corpus. Using our corpus, we train a NER model that achieves entity-level Precision of 72.81-93.98% with 58.14-81.77% Recall on a manually-annotated test dataset. Furthermore, we show that using a weighted loss function helps to combat class imbalance in token classification tasks. To make it easy for others to reproduce and build upon our work, we publicly release our corpus, models, and experimental code.
ER -
NOVOTNÝ, Vít, Kristýna LUGER, Michal ŠTEFÁNIK, Tereza VRABCOVÁ a Aleš HORÁK. People and Places of Historical Europe: Bootstrapping Annotation Pipeline and a New Corpus of Named Entities in Late Medieval Texts. Online. In Anna Rogers, Jordan Boyd-Graber, Naoaki Okazaki. \textit{Findings of the Association for Computational Linguistics: ACL 2023}. Toronto, Canada: Association for Computational Linguistics, 2023, s.~14104-14113. ISBN~978-1-959429-62-3.