Reconsidering SMT Over NMT for Closely Related Languages: A Case Study of Persian-Hindi Pair
This provides a practical solution for improving translation accuracy in specific language pairs where resources are limited, though it is incremental as it revisits older methods in a new context.
The paper tackles the problem of machine translation for closely related languages in moderate-resource settings, showing that Phrase-Based Statistical Machine Translation (PBSMT) outperforms Transformer-based Neural Machine Translation (NMT) with a BLEU score of 66.32 versus 53.7 for the Persian-Hindi pair.
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.