Semantic Prosody in Machine Translation: the English-Chinese Case of Passive Structures

2025-10-16

AI Summary

This study addresses the absence of semantic prosody modeling for passive constructions in machine translation, specifically focusing on the negative semantic prosody associated with Chinese *bèi*-passives in English-to-Chinese translation. We introduce the first semantic-prosody-annotated English–Chinese parallel dataset and propose a semantic prosody knowledge injection method, enabling the first explicit modeling of the negative prosody of *bèi*-passives. Fine-tuning on OPUS-MT, NLLB-600M, and mBART50 demonstrates that our approach significantly improves the accuracy of *bèi*-passive generation in negative contexts while effectively suppressing erroneous usage in neutral or positive contexts. Moreover, the injected prosodic knowledge exhibits cross-lingual transferability, generalizing to other multilingual models. This work fills a critical gap in semantic prosody modeling for machine translation and establishes a novel paradigm for prosody-aware neural machine translation.

Technology Category

• Natural Language Processing: Sentence-level Semantics, Textual Inference, etc.
• Reasoning under Uncertainty: Relational Probabilistic Models
• Machine Learning: Large Multimodal Models (LMMs)

Application Category

• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Search and Retrieval-Augmented AI: Multilingual and cross-lingual Web search
• Web Mining and Content Analysis: Mining multimedia, multimodal, multilingual, cross-lingual Web data

Abstract

Semantic prosody is a collocational meaning formed through the co-occurrence of a linguistic unit and a consistent series of collocates, which should be treated separately from semantic meaning. Since words that are literal translations of each other may have different semantic prosody, more attention should be paid to this linguistic property to generate accurate translations. However, current machine translation models cannot handle this problem. To bridge the gap, we propose an approach to teach machine translation models about semantic prosody of a specific structure. We focus on Chinese BEI passives and create a dataset of English-Chinese sentence pairs with the purpose of demonstrating the negative semantic prosody of BEI passives. Then we fine-tune OPUS-MT, NLLB-600M and mBART50 models with our dataset for the English-Chinese translation task. Our results show that fine-tuned MT models perform better on using BEI passives for translating unfavourable content and avoid using it for neutral and favourable content. Also, in NLLB-600M, which is a multilingual model, this knowledge of semantic prosody can be transferred from English-Chinese translation to other language pairs, such as Spanish-Chinese.

Problem

• Machine translation models ignore semantic prosody differences between languages
• Literal translations fail to convey negative connotations of Chinese BEI passives
• Current MT systems cannot handle collocational meaning in passive structures

Innovation

• Fine-tuned MT models handle semantic prosody
• Created dataset for negative BEI passive prosody
• Transferred prosody knowledge across language pairs