A Comparative Study of Translation Features of Cultural-loaded Terms in the 2025 Government Work Report—A Corpus-based Analysis of the Official Translation and AI Translation

This study constructs the Chinese-English Parallel Corpus of Cultural-Loaded Terms in the 2025 Government Work Report, abbreviated as 2025GWR-CLPC. From the corpus, 86 cultural-loaded terms are selected and divided into three categories, namely the policy and philosophy category, the system and mechanism category, and the livelihood and development category. Based on the Skopos Theory, this study adopts many methods including parallel corpus com-parison, quantitative and qualitative analysis, and audience comprehension assessment to compare the translation features of the above-mentioned terms in the official English version and the Doubao AI English version. The research results show that literal translation and free translation are the main strategies adopted by the two different versions. Official translation fits professional groups and formal occasions better, while AI translation is more ideal for the general public and quick information transmission. For improving the translation effect of culture-loaded expressions in political texts, it is necessary to combine the strengths of the two translation types, set up a standardized term base, and emphasize audience-focused translation.