AI and corpora for interpreting: Case study on ChatGPT-5.5 Instant and SCIPPC applied on glossary building

本论文以人工智能为研究对象，通过案例分析探讨人工智能在国际环境外交中是否能为口译员提供应用，尤其是关于术语提取与术语表构建。论文的结构分为三部分。第一部分，首先将对欧盟、美国与中国之间的国际关系进行简述。该国家之所以被选，是因为它们在外交领域中带来了极大的影响。其次，还将介绍口译员在外交场合下的角色，以及他们在进行翻译当中面临的哪些挑战。第二部分，将描述口译员能使用的一些术语管理手段，包括术语学、术语编纂、口译研究中最有影响力的理论，以及语料库与人工智能。第三部分，将分析比较两个术语表。首先，术语是从Speech Corpus of Interpreted Premier Press Conference语料库提取的。第一个分析对象是通过半自动提取方式与术语编纂构建的。第二个分析对象则是通过ChatGPT5.5 Instant构建的。最终目标是确认ChatGPT-5.5 Instant在术语提取与术语编纂当中的应用。分析的焦点在于术语提取的过程与术语表构建的质量。

This thesis focuses on a possible application of artificial intelligence in term mining and glossary building for diplomatic interpreting about the environment. It is divided into three chapters. The first chapter consists of a brief description of relationships between three of the current most important actors in international diplomacy, being the EU, the US and China, and an introduction to the role of the interpreter in this context as well as the challenges they face when providing their linguistic services. The second chapter introduces some of the tools interpreters can use to handle terminological management, focusing primarily on the most influential theories in interpreting studies, terminology and terminography, as well as corpora and AI. The third chapter consists of a comparative study concerning the application of ChatGPT-5.5 Instant in term mining and terminographical research on the Speech Corpus of Interpreted Premier Press Conference. Specifically, the analysis focused on the building process of two glossaries, one created through semiautomatic term mining and terminographical research and the other with the assistance of ChatGPT-5.5 Instant, as well as their quality.