Machine Translation

Unlike translation memory, machine translation (MT) actually translates text that is not in a database. Most MT systems grammatically analyze (parse) the source text and then generate a translation based on this analysis. Newer technologies are based on statistical analysis or on a hybrid approach that leverages both linguistic and statistical analysis.

The results of MT are generally not as good as translations produced by humans, but are useful for understanding roughly what a text says (a process known as “gisting”). MT is often used to determine whether a human translator should prepare a high-quality translation. Through Internet translation portals that make use of machine translation, it is perhaps the most commonly used translation tool in the world, but such results are seldom incorporated into localized documents for publication. In some cases, human translators edit machine translation results to produce final translations in what is called “post-editing.”

Machine translation can be used to provide very high-quality translations under certain controlled circumstances. It is also critical for providing translations of materials (such as results from database queries) that are time-sensitive and which cannot wait for time required for human translation. (To learn more about machine translation, please see the LISA Best Practice Guide, Implementing Machine Translation.)