Do not confuse Translation Memory (TM) with Machine Translation (MT)

Translation Memory or TM for short is basically a database that stores “segments”, which can be sentences, paragraphs, terms or sentence-like units that have previously been translated, in order to aid human translators. It is a technology that enables translators to store and reuse both the source and target language content for any translation project.

To fully realize the benefits of Translation Memory, translation memory tools must be used to create, manage and maintain multilingual glossaries along with the translation memories. Glossaries and translation memories are important intellectual properties that are usually delivered to clients along with the translated documents.

The benefits of Translation Memory: 

1. It enables consistency in translations on larger projects, where many translator are deployed to work on the project.
2. It improves productivity and accuracy of translation.
3. It enables cost savings for your translation projects through “leveraged” pricing on exact matches, repetitive text, fuzzy match text, etc.
4. It enables ongoing or future updates of your translation projects to be handled in the most cost-effective and time-efficient manner.

However, we should not confuse Translation Memory (TM) with Machine Translation (MT). Translation Memory (TM)  is not Machine Translation (MT). Machine translation (MT) is a software that produces very raw, draft translations automatically, such as those provided online by Google Translate, ToggleText, Systran, etc.

What Machine Translation does is that it performs simple substitution of words in one language for words in another language. As such, it usually cannot yield a good translation of a text because recognition of whole phrases and their closest counterparts in the target language is needed.