Is there a nice solution to the long-standing issue associated with the recognition (or displaying) of word order differences, except for intentionally lowering the match rate?
(In the example below, the positions of "Buyer" and "Seller" are different, but these differecnes are not highlighted)
Igor told us that the fuzzy matching algorithm doesn't take word order into account, so, to answer your question: no, I'm afraid.
>> Igor told us that the fuzzy matching algorithm doesn't take word order into account, ...
If so, this must be a challenge to be overcome, I think.
>If so, this must be a challenge to be overcome, I think.
If you're lucky (and your source and target syntax isn't too different), you can use QA glossary consistency to discover errors in segments like these. This probably won't work for JP <> EN ...
Your second guess is right. QA is useless in cases like this.
So it's not possible to define the JP equivalents of: