arXiv cs.CL· Telem Joyson Singh, Sanabam Boyshana, Sanasam Ranbir Singh, Priyankoo Sarmah·· 13 小时前AI 评分18
首个英语-米佐语法律机器翻译数据集发布：13,168 句对与 NLLB 基准测试
Towards English - Mizo Legal Machine Translation: Corpus Creation, Analysis, and Benchmarks
AI 导读
研究团队构建了首个英语-米佐语印度法院判决平行数据集，含来自高哈蒂高等法院和印度最高法院的 13,168 个句对，由米佐语母语专业译员翻译并经人工对齐和双人复核。在 1,000 句对的平衡法律测试集上微调 NLLB 200 1.3B 模型后达到 27.47 BLEU，比基座模型提升 7.16 BLEU，但使用英文词汇的频率仍约为人工译员的三倍。
正文 · 原文
Abstract:Indian court proceedings are held in English, so speakers of languages with limited translation technology have limited access to them. We present the first English Mizo parallel dataset of Indian court judgments. It contains 13,168 sentence pairs from the Gauhati High Court and the Supreme Court of India. Professional native Mizo translators translated complete documents; sentence pairs were manually aligned, and two independent reviewers checked the results. Our study finds that English words are borrowed in a very selective way: official and legal terms are almost always kept in English, but words for everyday things like actions, situations, and results are usually expressed in native Mizo. Mizo diacritics are also used inconsistently across the corpus. When we fine tune the NLLB 200 1.3B model on our training data, it scores 27.47 BLEU on a balanced legal test set of 1,000 sentence pairs, an improvement of 7.16 BLEU over the base model. However, it still uses English words about three times more often than a human translator does.
| Comments: | Accepted in OCOCOSDA 2026 |
| Subjects: | Computation and Language (cs.CL) |
| Cite as: | arXiv:2610.29943 [cs.CL] |
| (or arXiv:2610.29943v1 [cs.CL] for this version) | |
| https://doi.org/10.48550/arXiv.2610.29943 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Priyankoo Sarmah [view email]
[v1]
Wed, 21 Oct 2026 16:11:01 UTC (166 KB)
来源：arXiv cs.CL · arxiv.org