Speaker
Description
The present study aims to critically evaluate the performance of the most popular AI translation tools in accurately translating culturally rich Chinese texts. Specifically, it seeks to assess the quality of translations regarding cultural sensitivity, idiomatic accuracy, and contextual appropriateness. The methodology followed includes a comparative analysis conducted using a curated corpus of Chinese texts encompassing various genres— classic literature, recent news texts from several different domains (society, culture, technology and science) from the platform Chairman’s Bao, contemporary colloquial conversations from Mandarin Bean and specialised technical manuals accessible through Baidu Wenku. These platforms have been selected for their comprehensive representation of linguistic and cultural diversity. This diverse collection ensures a robust test of AI translation capabilities across different text genres. For the analysis, several leading AI translation tools will be evaluated. The tools will be assessed using both qualitative and quantitative methods. The study is expected to find varied effectiveness among the AI tools, with anticipated deficiencies in translating idiomatic expressions and maintaining cultural authenticity. The findings will offer insights into the current capabilities and limitations of AI translation tools, emphasising areas where AI can augment human translation efforts and identifying gaps that require further technological refinement with human contribution.
| Presenting author | Iulia Cîndea, PhD |
|---|
