Translation memories enable translation to be stored for future references. But how do you figure out how to use these stored translations? Find out how TM matches help translators look up similar strings to save up time, keep their translation consistency, and increase efficiency in their work.
What is a TM Match? TM Matches describe how closely the string being translated is similar to the ones already stored in the translation memory.
When translating a new document, each string would be compared to the translations stored in the Translation Memory, and TM Matches would show the similarity in percentages – from 100% match to a 0% match. If the string being translated has the exact same string translation in the TM, the TM Match would show 100% match. But if there is no similarity in the string and the ones stored in the TM, it would show a 0% match.
To give you an example, let’s say that there are 4 words in a string. If all of these words have been translated previously and is already stored in the translation memory, the TM Match would be 100%. This means that word by word, the string has already been translated and the TM has the same copy, in the exact same way.
If the last two words of the string are different from the ones stored in the translation memory, the TM Match would only show 50%. Since the last two words are not similar, but the other half, which is 50%, has already be translated and stored in the translation memory.
If there is nothing similar in the string – meaning that the string encountered is unique or new – the TM Match would show a 0% match, to represent that there is no stored translation that is similar to the string being translated.
Typically, matches range from 0% to 100%, depending on how the content are similar. But there is also a special case of a 101% match, which will be discussed in the next articles.