A recent study led by Li Zi from the School of Foreign Languages at the University of Sanya has unveiled a groundbreaking approach to translating the historical literature of Geng Lu Bu, a significant piece of China’s intangible cultural heritage. This research, published in the journal Applied Mathematics and Nonlinear Sciences, harnesses the power of multimodal machine translation, an innovative technology that combines textual and visual data to enhance understanding and accuracy in translation.
Geng Lu Bu embodies the spirit of maritime exploration and the legacy of seafaring practices in China, making its literature crucial for understanding the country’s historical relationship with the sea. The study introduces a multimodal network that selects and integrates visual information with textual features, allowing for a more nuanced translation process. As Li Zi explains, “The generated image features and text features are deeply interconnected at multiple levels and granularities.” This deep integration not only boosts the quality of translations but also preserves the cultural essence embedded in the original texts.
The research involved conducting translation comparison experiments using two datasets, VATEX and MSVD-Turkish, achieving impressive BLEU scores of 37.43 and 37.72. These metrics are a standard measure for evaluating the quality of machine-generated translations, indicating that the proposed model significantly outperforms traditional methods. The study further explores three different translation techniques—multimodal machine translation, human translation, and traditional translation—highlighting the effectiveness of the multimodal approach. Results show that this model produces translations with a high degree of text complexity and formality, offering a well-rounded and aesthetically pleasing sentence style.
For maritime professionals, the implications of this research are particularly exciting. The ability to accurately translate historical maritime literature can enhance cultural exchange and education, opening new avenues for tourism and heritage preservation. Companies involved in maritime education or tourism could leverage this technology to provide richer experiences for visitors interested in maritime history. Furthermore, the fusion of visual and textual data may also be applied in training simulations, where understanding historical contexts can inform modern maritime practices.
As the maritime sector continues to globalize, tools like the multimodal machine translation system developed by Li Zi could become invaluable. They not only facilitate better communication across language barriers but also foster a deeper appreciation of cultural heritage within the maritime community. This research stands as a testament to how technology can bridge the past and present, enhancing our understanding of the sea and its rich history.
In summary, the innovative approach presented in Li Zi’s study not only pushes the boundaries of translation technology but also opens up commercial opportunities for the maritime sector, enriching interactions and fostering a greater appreciation for maritime heritage. The findings, published in Applied Mathematics and Nonlinear Sciences, set a promising precedent for future research and applications in this vital field.
