Machine Translation Post-Editing

Machine translation has certainly been making huge leaps lately. While the technology has been around and running for almost 70 years now, there are reasons to believe that with progress made in the AI technology from major players like Google, Facebook, and LinkedIn, things may speed up significantly in the next few years.

Machine translation post-editing is the perfect combination of man and machine. The ideal process is to “train” the machine (i.e. “feed” it with accurate data) before starting the project. The better the data, the better the output. This MT output will then be edited and proofread [links] by specially trained linguists whose skill set is very different from regular translators and editors. Ideally, this step eliminates all machine-made errors while at the same time increasing the readability and style to TEP-quality and lowering the costs and increasing the volume.