Evaluation of Artificial Intelligence Translation
摘要
The proliferation of technological innovations has precipitated the rapid diffusion of information, fundamentally reshaping modes of communication. Language and translation technologies, which have exerted considerable influence over societal dynamics, are no anomaly in this paradigm shift. The paper investigates the use of ChatGPT, an online statistical machine translation tool, for translating specialised German texts into Slovak, focusing on ten articles from the German weekly business news magazine WirtschaftsWoche related to various aspects of the stock exchange. It outlines the challenges encountered when employing automated translation systems for these texts. The translated texts undergo quantitative and qualitative analyses to assess error rate, efficacy, and fidelity to the original context. Challenges include linguistic complexities, domain-specific terminologies, and contextual nuances inherent in the stock exchange field. Through rigorous evaluation methods, including error rate calculations and qualitative assessments, the paper aims to prove insights into the effectiveness of automated translation systems for specialised text genres. It contributes to the discourse on language technology, specialised domains, and translation studies, and provides information about future developments in machine translation research and its practical applications.