@inproceedings{zhang-etal-2026-language-models,
title = "Language Models for the Restoration of {L}atin Legal Manuscripts",
author = "Zhang, Shibingfeng and
Caraffa, Edoardo and
Zuffrano, Annafelicia and
Modesti, Maddalena and
Colavizza, Giovanni",
editor = "Sprugnoli, Rachele and
Passarotti, Marco",
booktitle = "Proceedings of the Fourth Workshop on Language Technologies for Historical and Ancient Languages ({LT}4{HALA} 2026) @ {LREC} 2026",
month = may,
year = "2026",
address = "Palma, Mallorca (Spain)",
publisher = "ELRA Language Resources Association (ELRA)",
url = "https://aclanthology.org/2026.lt4hala-1.37/",
doi = "10.63317/5g8oqsnzp537",
pages = "361--367",
abstract = "The collection of historical notarial documentation from Bologna is a valuable source, providing deep insights into the city{'}s institutional, legal, and socio-economic history. However, many of these manuscripts have sustained physical damage during centuries of conservation, rendering the text incomplete. To address this, we explored the restoration of these Latin notary documents using encoder-based pre-trained language models (PLMs) under the assumption that the length of missing text is known by estimation from the physical damage. We address the structural misalignment between the physical lacuna of the manuscript and the subword tokenization schemes of PLMs by designing an iterative decoding strategy to align model predictions with the known physical dimensions of lacuna. We also compared the efficacy of monolingual versus multilingual pre-training. Our strategy significantly outperforms baselines consist of standard decoding methods. Furthermore, stratified analysis across different text sections reveals that while monolingual models achieve better performance in general, multilingual models show a suggestive advantage in lexically dense segments, though this finding is not statistically significant. Overall, the best performance achieved by our method is a Hit@1 rate of 35.47{\%} in the short-span setting and 18.75{\%} in the long-span setting. While fully autonomous restoration remains an open challenge, our system provides a useful assistive tool for paleographers."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="zhang-etal-2026-language-models">
<titleInfo>
<title>Language Models for the Restoration of Latin Legal Manuscripts</title>
</titleInfo>
<name type="personal">
<namePart type="given">Shibingfeng</namePart>
<namePart type="family">Zhang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Edoardo</namePart>
<namePart type="family">Caraffa</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Annafelicia</namePart>
<namePart type="family">Zuffrano</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Maddalena</namePart>
<namePart type="family">Modesti</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Giovanni</namePart>
<namePart type="family">Colavizza</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 Fourth Workshop on Language Technologies for Historical and Ancient Languages (LT4HALA 2026) @ LREC 2026</title>
</titleInfo>
<name type="personal">
<namePart type="given">Rachele</namePart>
<namePart type="family">Sprugnoli</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Marco</namePart>
<namePart type="family">Passarotti</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>The collection of historical notarial documentation from Bologna is a valuable source, providing deep insights into the city’s institutional, legal, and socio-economic history. However, many of these manuscripts have sustained physical damage during centuries of conservation, rendering the text incomplete. To address this, we explored the restoration of these Latin notary documents using encoder-based pre-trained language models (PLMs) under the assumption that the length of missing text is known by estimation from the physical damage. We address the structural misalignment between the physical lacuna of the manuscript and the subword tokenization schemes of PLMs by designing an iterative decoding strategy to align model predictions with the known physical dimensions of lacuna. We also compared the efficacy of monolingual versus multilingual pre-training. Our strategy significantly outperforms baselines consist of standard decoding methods. Furthermore, stratified analysis across different text sections reveals that while monolingual models achieve better performance in general, multilingual models show a suggestive advantage in lexically dense segments, though this finding is not statistically significant. Overall, the best performance achieved by our method is a Hit@1 rate of 35.47% in the short-span setting and 18.75% in the long-span setting. While fully autonomous restoration remains an open challenge, our system provides a useful assistive tool for paleographers.</abstract>
<identifier type="citekey">zhang-etal-2026-language-models</identifier>
<identifier type="doi">10.63317/5g8oqsnzp537</identifier>
<location>
<url>https://aclanthology.org/2026.lt4hala-1.37/</url>
</location>
<part>
<date>2026-05</date>
<extent unit="page">
<start>361</start>
<end>367</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Language Models for the Restoration of Latin Legal Manuscripts
%A Zhang, Shibingfeng
%A Caraffa, Edoardo
%A Zuffrano, Annafelicia
%A Modesti, Maddalena
%A Colavizza, Giovanni
%Y Sprugnoli, Rachele
%Y Passarotti, Marco
%S Proceedings of the Fourth Workshop on Language Technologies for Historical and Ancient Languages (LT4HALA 2026) @ LREC 2026
%D 2026
%8 May
%I ELRA Language Resources Association (ELRA)
%C Palma, Mallorca (Spain)
%F zhang-etal-2026-language-models
%X The collection of historical notarial documentation from Bologna is a valuable source, providing deep insights into the city’s institutional, legal, and socio-economic history. However, many of these manuscripts have sustained physical damage during centuries of conservation, rendering the text incomplete. To address this, we explored the restoration of these Latin notary documents using encoder-based pre-trained language models (PLMs) under the assumption that the length of missing text is known by estimation from the physical damage. We address the structural misalignment between the physical lacuna of the manuscript and the subword tokenization schemes of PLMs by designing an iterative decoding strategy to align model predictions with the known physical dimensions of lacuna. We also compared the efficacy of monolingual versus multilingual pre-training. Our strategy significantly outperforms baselines consist of standard decoding methods. Furthermore, stratified analysis across different text sections reveals that while monolingual models achieve better performance in general, multilingual models show a suggestive advantage in lexically dense segments, though this finding is not statistically significant. Overall, the best performance achieved by our method is a Hit@1 rate of 35.47% in the short-span setting and 18.75% in the long-span setting. While fully autonomous restoration remains an open challenge, our system provides a useful assistive tool for paleographers.
%R 10.63317/5g8oqsnzp537
%U https://aclanthology.org/2026.lt4hala-1.37/
%U https://doi.org/10.63317/5g8oqsnzp537
%P 361-367
Markdown (Informal)
[Language Models for the Restoration of Latin Legal Manuscripts](https://aclanthology.org/2026.lt4hala-1.37/) (Zhang et al., LT4HALA 2026)
ACL
- Shibingfeng Zhang, Edoardo Caraffa, Annafelicia Zuffrano, Maddalena Modesti, and Giovanni Colavizza. 2026. Language Models for the Restoration of Latin Legal Manuscripts. In Proceedings of the Fourth Workshop on Language Technologies for Historical and Ancient Languages (LT4HALA 2026) @ LREC 2026, pages 361–367, Palma, Mallorca (Spain). ELRA Language Resources Association (ELRA).