Post-Editing of Machine Translation: Creating a Better Translation of Cultural Specific Terms

Abstract

The knowledge of machine translation has great importance in the pre-translation process, particularly when translating a literary work. There were a plethora of studies about the translation of cultural-specific items (CSI) with machines. However, they have not addressed machine translation post-editing to improve its translation. This study aims at describing the semantic analysis in the translation of cultural-specific items (CSI) by machine translation (MT) from Indonesian into English, analyzing their translation and whether or not they have similar meanings between the source text and target text to get a better quality translation. Qualitative research using descriptive methods was applied. The data were 12 sentences of CSI in the form of concrete and five sentences of CSI regarding socio-cultural terms from the short story of Betawi folklore entitled Angan-Angan si Muin (Muin's Wishful Thinking). The result revealed that the 2nd generated the better translation in concrete. Both machine translations (MT) produced different translations for translating CSI of socio-cultural terms. Therefore, post-editing is taken to improve their translation with semantic analysis of relation meaning with primary considerations such as material, shape, size, function, and description. Machines translator (MT) helps us to work efficiently and improve accuracy. But, the results of machine translations must still be reviewed by human translators to ensure key quality requirements are satisfied, including fidelity or correctness, intelligibility or clarity, and style, so we need post-editing translation.