Machine translation (MT) with post-editing (PE) is fast becoming a standard practice in our industry. This means that organizations need to be able to easily identify, qualify, train and evaluate post-editors’ performances.
Today, there are many methodologies in use, resulting in a lack of cohesive standards as organizations take various approaches for evaluating performance. Some use final output quality evaluation or post-editor productivity as a standalone metric. Others analyze quality data such as “over-edit” or “under-edit” of the post-editor’s effort or evaluate the percentage of MT suggestions used versus MT suggestions that are discarded in the final output.