A COMPARATIVE ANALYSIS OF TRANSLATION TECHNIQUES AND READABILITY OF ENG-IND FOOTBALL NEWS HEADLINES TRANSLATIONS GENERATED BY CHATGPT AND DEEPSEEK

The rapid development of artificial intelligence (AI) technology has brought significant changes to various aspects, including translation. AI systems such as ChatGPT and DeepSeek are now capable of translating various linguistic units, including headlines particularly in sports news headlines such as football, where the language is typically concise, dense, and often incorporates idiomatic expressions and metaphors. However, the quality of these translations specifically their readability, still needs to be evaluated. This study aims to analyze translation techniques according to Molina & Albir (2002), and to evaluate and compare the readability of football news headlines translations generated by ChatGPT and DeepSeek according to Nababan et al. (2012). This study employs a qualitative descriptive method with a comparative approach. The data used consists of 100 football news headlines from the bbc.com. The results show that the most frequently used translation technique by ChatGPT is amplification (38.5%), while DeepSeek’s is borrowing (30%). In terms of readability, ChatGPT outperforms DeepSeek with an average score of 2.26 compared to DeepSeek’s average score of 2.21.