Doc-Guided Sent2Sent++: A Sent2Sent++ Agent with Doc-Guided memory for Document-level Machine Translation
Fuente:
arXiv
Saved in:
| Main Authors: | , , , , , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866909456963469312 |
|---|---|
| author | Guo, Jiaxin Luo, Yuanchang Wei, Daimeng Zhang, Ling Li, Zongyao Shang, Hengchao Rao, Zhiqiang Li, Shaojun Yang, Jinlong Wu, Zhanglin Yang, Hao |
| author_facet | Guo, Jiaxin Luo, Yuanchang Wei, Daimeng Zhang, Ling Li, Zongyao Shang, Hengchao Rao, Zhiqiang Li, Shaojun Yang, Jinlong Wu, Zhanglin Yang, Hao |
| contents | The field of artificial intelligence has witnessed significant advancements in natural language processing, largely attributed to the capabilities of Large Language Models (LLMs). These models form the backbone of Agents designed to address long-context dependencies, particularly in Document-level Machine Translation (DocMT). DocMT presents unique challenges, with quality, consistency, and fluency being the key metrics for evaluation. Existing approaches, such as Doc2Doc and Doc2Sent, either omit sentences or compromise fluency. This paper introduces Doc-Guided Sent2Sent++, an Agent that employs an incremental sentence-level forced decoding strategy \textbf{to ensure every sentence is translated while enhancing the fluency of adjacent sentences.} Our Agent leverages a Doc-Guided Memory, focusing solely on the summary and its translation, which we find to be an efficient approach to maintaining consistency. Through extensive testing across multiple languages and domains, we demonstrate that Sent2Sent++ outperforms other methods in terms of quality, consistency, and fluency. The results indicate that, our approach has achieved significant improvements in metrics such as s-COMET, d-COMET, LTCR-$1_f$, and document-level perplexity (d-ppl). The contributions of this paper include a detailed analysis of current DocMT research, the introduction of the Sent2Sent++ decoding method, the Doc-Guided Memory mechanism, and validation of its effectiveness across languages and domains. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2501_08523 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Doc-Guided Sent2Sent++: A Sent2Sent++ Agent with Doc-Guided memory for Document-level Machine Translation Guo, Jiaxin Luo, Yuanchang Wei, Daimeng Zhang, Ling Li, Zongyao Shang, Hengchao Rao, Zhiqiang Li, Shaojun Yang, Jinlong Wu, Zhanglin Yang, Hao Computation and Language Artificial Intelligence The field of artificial intelligence has witnessed significant advancements in natural language processing, largely attributed to the capabilities of Large Language Models (LLMs). These models form the backbone of Agents designed to address long-context dependencies, particularly in Document-level Machine Translation (DocMT). DocMT presents unique challenges, with quality, consistency, and fluency being the key metrics for evaluation. Existing approaches, such as Doc2Doc and Doc2Sent, either omit sentences or compromise fluency. This paper introduces Doc-Guided Sent2Sent++, an Agent that employs an incremental sentence-level forced decoding strategy \textbf{to ensure every sentence is translated while enhancing the fluency of adjacent sentences.} Our Agent leverages a Doc-Guided Memory, focusing solely on the summary and its translation, which we find to be an efficient approach to maintaining consistency. Through extensive testing across multiple languages and domains, we demonstrate that Sent2Sent++ outperforms other methods in terms of quality, consistency, and fluency. The results indicate that, our approach has achieved significant improvements in metrics such as s-COMET, d-COMET, LTCR-$1_f$, and document-level perplexity (d-ppl). The contributions of this paper include a detailed analysis of current DocMT research, the introduction of the Sent2Sent++ decoding method, the Doc-Guided Memory mechanism, and validation of its effectiveness across languages and domains. |
| title | Doc-Guided Sent2Sent++: A Sent2Sent++ Agent with Doc-Guided memory for Document-level Machine Translation |
| topic | Computation and Language Artificial Intelligence |
| url | https://arxiv.org/abs/2501.08523 |