在波斯语-印地语这对相似语言上，统计机器翻译反而优于神经网络翻译。
Reconsidering SMT Over NMT for Closely Related Languages: A Case Study of Persian-Hindi Pair
- 用短语的统计方法翻译波斯语和印地语，效果更好。
- 统计方法得分为66.32，远超神经网络的53.7。
- 适合资源中等、语言结构相近的翻译任务。
本文表明，在中等资源条件下，对于结构相似的语言对（如波斯语-印地语），基于短语的统计机器翻译（PBSMT）可超越基于Transformer的神经机器翻译（NMT）。尽管Transformer通常依赖大量双语语料，但实验显示，PBSMT在相同数据集上取得66.32的BLEU分数，显著高于Transformer-NMT的53.7。我们还探索了SMT架构的变体，包括使用罗马化文本训练以及将波斯语句序调整为左到右（LTR）结构以匹配印地语。研究结果强调应根据语言对特性选择合适架构，主张即使在普遍由NMT主导的场景下，SMT仍是高效备选方案。
原文摘要 · Abstract (English)
This paper demonstrates that Phrase-Based Statistical Machine Translation (PBSMT) can outperform Transformer-based Neural Machine Translation (NMT) in moderate-resource scenarios, specifically for structurally similar languages, like the Persian-Hindi pair. Despite the Transformer architecture's typical preference for large parallel corpora, our results show that PBSMT achieves a BLEU score of 66.32, significantly exceeding the Transformer-NMT score of 53.7 on the same dataset. Additionally, we explore variations of the SMT architecture, including training on Romanized text and modifying the word order of Persian sentences to match the left-to-right (LTR) structure of Hindi. Our findings highlight the importance of choosing the right architecture based on language pair characteristics and advocate for SMT as a high-performing alternative, even in contexts commonly dominated by NMT.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。