Machine translation focuses mainly on high-resource languages (HRLs), while
low-resource languages (LRLs) like Taiwanese Hokkien are relatively
under-explored. This study aims to address this gap by developing a dual
translation model between Taiwanese Hokkien and both Traditional Mandarin
Chinese and English. We employ a pre-trained LLaMA2-7B model specialized in
Traditional Mandarin Chinese to leverage the orthographic similarities between
Taiwanese Hokkien Han and Traditional Mandarin Chinese. Our comprehensive
experiments involve translation tasks across various writing systems of
Taiwanese Hokkien and between Taiwanese Hokkien and other HRLs. We find that
the use of a limited monolingual corpus also further improve the model's
Taiwanese Hokkien capabilities. We then utilize our translation model to
standardize all Taiwanese Hokkien writing systems into Hokkien Han, resulting
in further performance improvements. Additionally, we introduce an evaluation
method incorporating back-translation and GPT-4 to ensure reliable translation
quality assessment even for LRLs. The study contributes to narrowing the
resource gap for Taiwanese Hokkien and empirically investigates the advantages
and limitations of pre-training and fine-tuning based on LLaMA 2.

本研究旨在填补机器翻译对台湾闽南语等低资源语言的相对忽视之处，通过开发台湾闽南语与繁体中文及英文的双向翻译模型，并利用预训练的 LLaMA2-7B 模型，综合实验跨越台湾闽南语的不同书写系统以及与其他高资源语言之间的翻译任务，最终大大提升了模型的翻译能力，并通过标准化所有台湾闽南语书写系统为福建汉字，进一步提高了性能；另外，引入反向翻译和 GPT-4 结合的评估方法，确保了对低资源语言的可靠翻译质量评估；此研究有助于缩小台湾闽南语的资源差距，并从实证角度研究了基于 LLaMA 2 的预训练和微调的优势和局限性。

探索和标准化四种写作系统，提升福建双语翻译

Enhancing Hokkien Dual Translation by Exploring and Standardizing of  Four Writing Systems

Taiwanese Hokkien is declining in use and status due to a language shift
towards Mandarin in Taiwan. This is partly why it is a low resource language in
NLP and speech research today. To ensure that the state of the art in speech
processing does not leave Taiwanese Hokkien behind, we contribute a 1.5-hour
dataset of Taiwanese Hokkien to ML-SUPERB's hidden set. Evaluating ML-SUPERB's
suite of self-supervised learning (SSL) speech representations on our dataset,
we find that model size does not consistently determine performance. In fact,
certain smaller models outperform larger ones. Furthermore, linguistic
alignment between pretraining data and the target language plays a crucial
role.

台語的使用和地位因台灣向普通話的語言轉變而下降。這是目前自然語言處理和語音研究中它成為低資源語言的部分原因。為了確保台語不落後於語音處理技術的最新發展，我們向 ML-SUPERB 的隱藏集合中貢獻了一個 1.5 小時的台語資料集。在我們的資料集上評估 ML-SUPERB 的自我監督學習（SSL）語音表示套件，發現模型大小並不一致地決定性能。實際上，某些較小的模型表現更好。此外，預訓練數據和目標語言之間的語言對齊起關鍵作用。