What conditions need to be fulfilled before we can have a context or 101% match? Find out how the translation memory takes past information and uses it to speed up the localization process.
What is a context or 101% match? And why is it important?
So, let’s say we’re translating this new sentence, which is AAA. When we have an already, we have this nice TM previously, and by any chance, it happens that it already contains the same sentence AAA and it’s translated as XXX. So because the sentences are the same, this is 100% match, right? It matches word by word. Now, what makes it the 101% match is the context of the sentence.
Now, how is the context determined? Context is simply determined by the preceding segment and the following segments. Okay. So what is the text before and what is the text after from the segment that we’re currently translating? And because the translation memory also stores this information of the context, so for the AAA that we already have in the translation memory, we have the context. And in case there is a match between the context of what we’re translating right now, and what we translated before, and we already established that the current sentence that we’re translating is the same, this is what would make this match a 101% match. Because the context is the same.
Now, why are 101% matches important, is because they give us more confidence that the translation from the TM will be accurate, because the context is the same. Okay, so it’s much stronger than 100% match only. And in this case, you might completely skip review of a 101% match again, because you have super confidence, because the context is the same. Now, what actually context and how context can affect the meaning of the word or the sentence, especially when we have something short, like let’s say simply home, and home? In the first example, let’s say this is preceded by, “Where are you? Home. Go to bed.” That’s example number one. Example number two, the preceding string is about us. And login, maybe. Okay. So while the sentence that we’re translating is the same, based on the context, we know that in example, number one, it’s referring to the home where you live with your family. And in case number not true, it is referring to home as a homepage. So this is how the context helps. And again, if we have the same context, as we had previously in TM, then we have a 101% match and we can be more confident that translation is accurate and maybe doesn’t even need to be reviewed by a person and we can just automatically apply translation from TM and not even bother, whether it’s correct or not.