The term post-editing or post-revision refers to the linguistic verification of texts produced by a machine translation engine (MT).
Depending on the objective of the translation project, the post-reviser may intervene more or less deeply on the text to achieve the level of quality desired by the customer and previously agreed upon with them, referred to as “high-quality human translation and revision”, or “publishable quality”, a level for which a full post-editing is recommended, or else, at a lower level, “sufficiently good” or “fit for the purpose” quality, in which case a lighter post-editing is recommended.
“Sufficiently good” refers to the level of quality that defines the final text as understandable (the main content of the text is understandable) and faithful (the final text conveys the same meaning as the original text), even if it is not stylistically convincing.
“High-quality human translation and revision” or “publishable quality” refers to the level of quality that defines the text as understandable (the end user perfectly understands the content of the text), faithful (the final text conveys the same meaning as the original text), and stylistically appropriate, even if the style may not match that of a text produced by a mother tongue human translator.
One of the fields in which post-editing is most widely used is that of technical-scientific translation: texts aimed at conveying sectoral messages with highly technical content, where the predominance of terms with univocal meaning minimizes the possibility of misunderstanding or non-contextual translation, indeed require less stylistic creativity and are thus the ideal ground for Machine Translation (MT) and, consequently, post-editing.