Large Language Models for Classical Chinese Poetry Translation: Benchmarking, Evaluating, and Improving
This work addresses the problem of high-quality translation of culturally rich classical Chinese poetry for researchers and practitioners in NLP, representing an incremental improvement through a domain-specific method.
The paper tackles the challenge of translating classical Chinese poetry, which requires both accuracy and poetic elegance, by introducing a benchmark (PoetMT) and a GPT-4-based metric to evaluate large language models (LLMs). It finds that existing LLMs fall short and proposes a Retrieval-Augmented Machine Translation (RAT) method that consistently outperforms comparison methods across multiple metrics, including BLEU, COMET, BLEURT, and human evaluation.
Different from the traditional translation tasks, classical Chinese poetry translation requires both adequacy and fluency in translating culturally and historically significant content and linguistic poetic elegance. Large language models (LLMs) with impressive multilingual capabilities may bring a ray of hope to achieve this extreme translation demand. This paper first introduces a suitable benchmark (PoetMT) where each Chinese poetry has a recognized elegant translation. Meanwhile, we propose a new metric based on GPT-4 to evaluate the extent to which current LLMs can meet these demands. Our empirical evaluation reveals that the existing LLMs fall short in the challenging task. Hence, we propose a Retrieval-Augmented Machine Translation (RAT) method which incorporates knowledge related to classical poetry for advancing the translation of Chinese Poetry in LLMs. Experimental results show that RAT consistently outperforms all comparison methods regarding wildly used BLEU, COMET, BLEURT, our proposed metric, and human evaluation.