Machine Translation Errors Identification and Pre-editing Strategies in Chinese to English Translation of Pharmaceutical Instructions

Abstract

With the increasing share of Chinese drugs in the international market, the translation quality of instructions has become an important factor affecting the internationalization of Chinese drugs. In recent years, machine translation has been widely used in the field of pharmaceutical instructions translation. However, its translation quality is sometimes unreliable and errors occur frequently, which brings potential risks to the accurate use and safety supervision of drugs. In this study, a total of 10 Chinese pharmaceutical instructions and their corresponding English translations were selected by convenient sampling method. Based on the Multidimensional Quality Metrics (MQM) model, this study classifies and analyzes the common errors in the translation of Chinese pharmaceutical instructions via DeepL. The findings show that the errors of Chinese pharmaceutical instructions are mainly concentrated in three aspects: terminology, accuracy and fluency. In response to the above issues, this thesis proposes six pre-editing strategies: explanation, terminology standardization, annotation, word order adjustment, omission, and addition. It is hoped to provide insights into the optimization of machine translation workflows in medical contexts and contribute to improving the translation quality of Chinese pharmaceutical instructions to better meet the international market demand.