The ACCURACY OF CHATGPT IN TRANSLATING LINGUISTICS TEXT IN SCIENTIFIC JOURNALS
Abstract
AI-generated Machine Translation, such as Neural Machine Translation, has transformed the traditional role of translators. There is a viral language model called ChatGPT. ChatGPT is a conversational variation of Natural Language Processing (NLP) Generative Pretrained Transformer (GPT) models. This research is aimed to analyze the accuracy of ChatGPT in translating linguistics text in scientific journals. The study employs qualitative approach and includes into quality assessment. The data source of this research is a National journal named Ranah focused on language and linguistics, Volume 12 number 1-13 published in 2023. The data then analyzed using few theories. Snover’s theory in finding error rate (HTER) and translation quality index to measure translation accuracy by Schiaffino and Zearo. Through the analyzation result, it could be inferred that In the analysis of 2034 words using Snover's theory revealed specific categories of translation errors in ChatGPT with a total of 6% error, there were 15 insertion errors (0.73%), 22 deletion errors (1.32%), 72 substitution errors (3.54%), and 17 shifting errors (0.83%). With a total error of 6,4%, this brings ChatGPT's accuracy rate in translating linguistic scientific texts to 93,6%. Through the analyzation result, it could be inferred that ChatGPT successfully translates scientific text within excellent category .
References
Khoshafah, F. (2023). ChatGPT for Arabic-English Translation: Evaluating the Accuracy. Research Square. DOI: 10.21203/rs.3.rs-2814154/v1.
Işım, Ç., & Balcıoğlu, Y. S. (2023). CHATGPT: PERFORMANCE OF TRANSLATE. In 3rd International ACHARAKA Congress on Humanities and Social Sciences Proceedings Book (pp. 47-51).
Saldanha, G., and O’Brien, S. (2014). Research Methodologies in Translation Studies. Routledge. New York.
Schiaffino, R. and F. Zeano. (2002). The Measurement of Quality in Translation Using Statistical Methods. 43rd ATA Conference 2002. J.D. Edwards and Lionbridge Technologies, Inc. Atlanta: USA. Accessed on: 20 May 2023
Sugiyono. (2011). Metode Penelitian Kuantitatif, Kualitatif, dan Kombinasi (Mixed Methods). Bandung: Alfabeta.
Siu, Sai Cheong. (2023). ChatGPT and GPT-4 for Professional Translators: Exploring the Potential of Large Language Models in Translation. 10.2139/ssrn.4448091.
Snover, M., Dorr, B., Schwartz, R., Makhoul, J., Micciulla, L., Weischedel, R. (2006). A Study of Translation Error Rate With Targeted Human Annotation. LAMP-TR-126.
Suryani, N. Y., and Fitria, T. N. (2022). Error Analysis of Abstract Translation in Scientific Writing by Using Google Translate. JETAL: JOURNAL OF ENGLISH TEACHING & APPLIED LINGUISTICS. Vol. 3, No. 2. Pp. 33-40. https://doi.org/10.36655/jetal.v3i2.669
Windari, M., & Al Hafizh, M. (2021). Error Analysis on Students’ Translation News Item Text from English to Indonesia.
