Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Wang, Kuang-Da, Ding, Shuoyang, Yang, Chao-Han Huck, Hsieh, Ping-Chun, Peng, Wen-Chih, Lavrukhin, Vitaly, Ginsburg, Boris
Format: Preprint
Veröffentlicht: 2025
Schlagworte:
Online-Zugang:https://arxiv.org/abs/2509.17249
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866911168080117760
author Wang, Kuang-Da
Ding, Shuoyang
Yang, Chao-Han Huck
Hsieh, Ping-Chun
Peng, Wen-Chih
Lavrukhin, Vitaly
Ginsburg, Boris
author_facet Wang, Kuang-Da
Ding, Shuoyang
Yang, Chao-Han Huck
Hsieh, Ping-Chun
Peng, Wen-Chih
Lavrukhin, Vitaly
Ginsburg, Boris
contents Despite Large Language Models (LLMs) demonstrating superior translation performance and long-context capabilities, evaluation methodologies remain constrained to sentence-level assessment due to dataset limitations, token number restrictions in metrics, and rigid sentence boundary requirements. We introduce SEGALE, an evaluation scheme that extends existing automatic metrics to long-document translation by treating documents as continuous text and applying sentence segmentation and alignment methods. Our approach enables previously unattainable document-level evaluation, handling translations of arbitrary length generated with document-level prompts while accounting for under-/over-translations and varied sentence boundaries. Experiments show our scheme significantly outperforms existing long-form document evaluation schemes, while being comparable to evaluations performed with groundtruth sentence alignments. Additionally, we apply our scheme to book-length texts and newly demonstrate that many open-weight LLMs fail to effectively translate documents at their reported maximum context lengths.
format Preprint
id arxiv_https___arxiv_org_abs_2509_17249
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Extending Automatic Machine Translation Evaluation to Book-Length Documents
Wang, Kuang-Da
Ding, Shuoyang
Yang, Chao-Han Huck
Hsieh, Ping-Chun
Peng, Wen-Chih
Lavrukhin, Vitaly
Ginsburg, Boris
Computation and Language
Despite Large Language Models (LLMs) demonstrating superior translation performance and long-context capabilities, evaluation methodologies remain constrained to sentence-level assessment due to dataset limitations, token number restrictions in metrics, and rigid sentence boundary requirements. We introduce SEGALE, an evaluation scheme that extends existing automatic metrics to long-document translation by treating documents as continuous text and applying sentence segmentation and alignment methods. Our approach enables previously unattainable document-level evaluation, handling translations of arbitrary length generated with document-level prompts while accounting for under-/over-translations and varied sentence boundaries. Experiments show our scheme significantly outperforms existing long-form document evaluation schemes, while being comparable to evaluations performed with groundtruth sentence alignments. Additionally, we apply our scheme to book-length texts and newly demonstrate that many open-weight LLMs fail to effectively translate documents at their reported maximum context lengths.
title Extending Automatic Machine Translation Evaluation to Book-Length Documents
topic Computation and Language
url https://arxiv.org/abs/2509.17249