Post-editing and translation management

Post-editing and Translation Management
Translation Management
Translation management has grown in popularity, especially in large projects where budgeting and deadlines are a major priority. Translation management, on a basic level, involves the substitution of words from one language to another, although this is usually insufficient to produce a good text translation since whole phrase recognition against close phrase counterparts in target language is a major requirement (Somers, 2013). Therefore, TM has adopted statistical techniques as a way of enhancing translations, as well as isolating anomalies and handling typology differences. Current TM software enables profession or domain-based customization, which limits the scope of substitutions, allowed and, therefore, improves output, especially where formulaic or formal language is required. As such, it would be expected that TM in cases involving legal and official documents can produce more palatable output as compared to less standardized text and conversations. Output quality in TM can also be enhanced through human intervention, such as the unambiguous identification of words that can be considered proper names in the text, which enables more accurate translation (Somers, 2013). The TM process can be described in terms of decoding the source text’s meaning, as well as re-encoding in the target language supported by a complex cognitive operation.
Post-editing refers to the process through which machine-generated translations are improved with little manual intervention or labor, and is closely linked to the concept of pre-editing. During translation of text, it is possible to achieve best results through source text pre-editing, such as through controlled language principles, after which the machine output is post-edited (Somers, 2013). Post-editing is quite different from editing that refers to improvement of text generated by humans, also known in translation as revision. Text that has been post-edited could be revised afterwards in order to ensure that language choices are of the best quality. Post-editing generally involves correction of output from translation, specifically to ensure that the output is of high quality. The degree or extent of post-editing may differ, which is dictated by the translation management criteria, and may either be full post-editing or light post-editing. Light post-editing involves minimal human intervention and is mainly meant to ensure that the output is readable, in which case the post-editor expects that the client only needs the text for inbound reasons (Somers, 2013). Full post-editing, on the other hand, seeks to make text more stylistically appropriate by achieving improved quality of text for outbound and inbound purposes.
Somers, H. L. (2013). Computers and translation: A translator’s guide. Amsterdam: J. Benjamins