@inproceedings{bruckner-etal-2026-modeling,
title = "Modeling the Language of Holocaust Survivors' Testimony with Domain-Adapted Transformers",
author = {Br{\"u}ckner, Christopher and
Lehe{\v{c}}ka, Jan and
{\v{S}}vec, Jan and
Pecina, Pavel},
editor = "Anuradha, Isuri and
Wynne, Martin",
booktitle = "Proceedings of The Second Workshop on Holocaust Testimonies as Language Resources ({HTR}es)",
month = may,
year = "2026",
address = "Palma, Mallorca (Spain)",
publisher = "ELRA Language Resources Association (ELRA)",
url = "https://aclanthology.org/2026.htres-2.9/",
doi = "10.63317/2cvjzqrxsjks",
pages = "74--83",
abstract = "Documents related to the Holocaust increasingly move into the focus of Natural Language Processing research, including the digitization of written text, the automatic transcription of oral archives, and interpretive downstream tasks such as Named Entity Recognition. However, most modern language models are trained primarily on modern text, and thus struggle with historical language, historical entities, and domain-specific terminology. Furthermore, transcribed speech introduces challenges such as transcription errors, noise, filler words, and dialectal speech not often contained in textual datasets. We present XLM-RoBERTa-malach, a text encoder domain-adapted to oral testimonies of Holocaust survivors in seven languages. In addition to descriptions of the data acquisition via Automatic Speech Recognition, data augmentation via Machine Translation, and the continued pretraining of a state-of-the-art multilingual transformer, we evaluate the domain-adapted model on the Named Entity Recognition task. Experiments on this task show superior performance over the general-domain transformer in a multilingual domain-specific setting, including languages not seen during the domain adaptation."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="bruckner-etal-2026-modeling">
<titleInfo>
<title>Modeling the Language of Holocaust Survivors’ Testimony with Domain-Adapted Transformers</title>
</titleInfo>
<name type="personal">
<namePart type="given">Christopher</namePart>
<namePart type="family">Brückner</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jan</namePart>
<namePart type="family">Lehečka</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jan</namePart>
<namePart type="family">Švec</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Pavel</namePart>
<namePart type="family">Pecina</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2026-05</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Proceedings of The Second Workshop on Holocaust Testimonies as Language Resources (HTRes)</title>
</titleInfo>
<name type="personal">
<namePart type="given">Isuri</namePart>
<namePart type="family">Anuradha</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Martin</namePart>
<namePart type="family">Wynne</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>ELRA Language Resources Association (ELRA)</publisher>
<place>
<placeTerm type="text">Palma, Mallorca (Spain)</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
</relatedItem>
<abstract>Documents related to the Holocaust increasingly move into the focus of Natural Language Processing research, including the digitization of written text, the automatic transcription of oral archives, and interpretive downstream tasks such as Named Entity Recognition. However, most modern language models are trained primarily on modern text, and thus struggle with historical language, historical entities, and domain-specific terminology. Furthermore, transcribed speech introduces challenges such as transcription errors, noise, filler words, and dialectal speech not often contained in textual datasets. We present XLM-RoBERTa-malach, a text encoder domain-adapted to oral testimonies of Holocaust survivors in seven languages. In addition to descriptions of the data acquisition via Automatic Speech Recognition, data augmentation via Machine Translation, and the continued pretraining of a state-of-the-art multilingual transformer, we evaluate the domain-adapted model on the Named Entity Recognition task. Experiments on this task show superior performance over the general-domain transformer in a multilingual domain-specific setting, including languages not seen during the domain adaptation.</abstract>
<identifier type="citekey">bruckner-etal-2026-modeling</identifier>
<identifier type="doi">10.63317/2cvjzqrxsjks</identifier>
<location>
<url>https://aclanthology.org/2026.htres-2.9/</url>
</location>
<part>
<date>2026-05</date>
<extent unit="page">
<start>74</start>
<end>83</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Modeling the Language of Holocaust Survivors’ Testimony with Domain-Adapted Transformers
%A Brückner, Christopher
%A Lehečka, Jan
%A Švec, Jan
%A Pecina, Pavel
%Y Anuradha, Isuri
%Y Wynne, Martin
%S Proceedings of The Second Workshop on Holocaust Testimonies as Language Resources (HTRes)
%D 2026
%8 May
%I ELRA Language Resources Association (ELRA)
%C Palma, Mallorca (Spain)
%F bruckner-etal-2026-modeling
%X Documents related to the Holocaust increasingly move into the focus of Natural Language Processing research, including the digitization of written text, the automatic transcription of oral archives, and interpretive downstream tasks such as Named Entity Recognition. However, most modern language models are trained primarily on modern text, and thus struggle with historical language, historical entities, and domain-specific terminology. Furthermore, transcribed speech introduces challenges such as transcription errors, noise, filler words, and dialectal speech not often contained in textual datasets. We present XLM-RoBERTa-malach, a text encoder domain-adapted to oral testimonies of Holocaust survivors in seven languages. In addition to descriptions of the data acquisition via Automatic Speech Recognition, data augmentation via Machine Translation, and the continued pretraining of a state-of-the-art multilingual transformer, we evaluate the domain-adapted model on the Named Entity Recognition task. Experiments on this task show superior performance over the general-domain transformer in a multilingual domain-specific setting, including languages not seen during the domain adaptation.
%R 10.63317/2cvjzqrxsjks
%U https://aclanthology.org/2026.htres-2.9/
%U https://doi.org/10.63317/2cvjzqrxsjks
%P 74-83
Markdown (Informal)
[Modeling the Language of Holocaust Survivors’ Testimony with Domain-Adapted Transformers](https://aclanthology.org/2026.htres-2.9/) (Brückner et al., htres 2026)
ACL