An Empirical Study on Chinese Character Decomposition in Multiword Expression-Aware Neural Machine Translation

2025-12-17

AI Summary

Chinese multi-word expressions (MWEs) suffer from semantic distortion in neural machine translation (NMT) due to ambiguity, low frequency, and non-compositionality—challenges inadequately addressed by Western subword methods (e.g., BPE), which are ill-suited for logographic writing systems. Method: This work pioneers systematic integration of character-level structural decomposition (radical/stroke granularity) into MWE-aware translation modeling. We propose a character-decomposition-based embedding enhancement mechanism, jointly optimized with an MWE identification module, and introduce a fine-grained semantic alignment strategy within the Transformer architecture. Contribution/Results: On a dedicated Chinese–English MWE test set, our approach achieves a +2.3 BLEU improvement over strong baselines, significantly enhancing idiom fidelity and literal meaning preservation. It establishes a novel, interpretable, and scalable lexical modeling paradigm for logographic-language NMT, bridging a critical gap between morphological awareness and subword segmentation in ideographic scripts.

Technology Category

• Natural Language Processing: Lexical Semantics and Morphology
• Machine Learning: Mixture of Experts (MoE)
• Cognitive Modeling & Cognitive Systems: Neural Spike Coding

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
• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs

Abstract

Word meaning, representation, and interpretation play fundamental roles in natural language understanding (NLU), natural language processing (NLP), and natural language generation (NLG) tasks. Many of the inherent difficulties in these tasks stem from Multi-word Expressions (MWEs), which complicate the tasks by introducing ambiguity, idiomatic expressions, infrequent usage, and a wide range of variations. Significant effort and substantial progress have been made in addressing the challenging nature of MWEs in Western languages, particularly English. This progress is attributed in part to the well-established research communities and the abundant availability of computational resources. However, the same level of progress is not true for language families such as Chinese and closely related Asian languages, which continue to lag behind in this regard. While sub-word modelling has been successfully applied to many Western languages to address rare words improving phrase comprehension, and enhancing machine translation (MT) through techniques like byte-pair encoding (BPE), it cannot be applied directly to ideograph language scripts like Chinese. In this work, we conduct a systematic study of the Chinese character decomposition technology in the context of MWE-aware neural machine translation (NMT). Furthermore, we report experiments to examine how Chinese character decomposition technology contributes to the representation of the original meanings of Chinese words and characters, and how it can effectively address the challenges of translating MWEs.

Problem

• Addresses MWE challenges in Chinese NMT
• Explores Chinese character decomposition for meaning representation
• Improves translation of multi-word expressions in Chinese

Innovation

• Chinese character decomposition for MWE-aware NMT
• Systematic study of ideograph decomposition in translation
• Addressing MWE challenges through character-level representation