• State Suggested Answer
  • Replies 2 replies
  • Answers 1 answer
  • Subscribers 435 subscribers
  • Views 2678 views
  • Users 0 members are here
Options
Related

Studio does not show an existing 100% match for a segment but offers another, less relevant (70%) match

JAR Kaantajat

Hello!

I use the professional version of SDL Trados Studio 2014 and yesterday I ran the latest (SP2) update on it. When I use the fuzzy function to translate the documents, Studio stops at a segment and shows a 70% match for it, though I know a 100% match exists. When I use the concordance search to look for the existing translation, the search does not give me the exact translation but some irrelevant translations that include some of the searched words. When I split the searched segment and perform the search with only one word in it, the concordance search shows me the 100% match that I was looking for. I use the translation memory of the size of 84 000 kb (with 86 000 TU's) locally, so the problem does not have anything to do with slow servers, etc. I also recomputed the fuzzy index statistics, but it didn't help. I really can't imagine what is causing this problem, so could anyone help? Thanks in advance.

Best Regards,

Kati

on behalf of Jar Kaantajat
Rate translation
×
Suggest better translation
×
Moderator UI
×
  • Get AI Suggestion
  • 0
    Hi Kati/Jar,

    The only reason I can think for this is that there are differences in the matches and you are just not seeing them... maybe hidden tags or other minor differences. If you haven't managed to resolve this yet perhaps you can share some screenshots showing the 70% match in the TM Lookup and then the exact match found through concordance?

    Paul Filkin | RWS Group

    ________________________
    Design your own training!
    You've done the courses and still need to go a little further, or still not clear? 
    Tell us what you need in our Community Solutions Hub

  • 0 in reply to Paul
    Thanks for your reply, Paul. This problem can't be caused by hidden tags, as we know the tags and their places very well in the xml files we translate and we always see them (including their differences) in Studio. I haven't had this problem lately and I can't think of any reason for it. However, I constantly have other match problems with the fuzzy function. Fuzzy very often stops at segments that have two 100% matches and the translation results window therefore shows one of them as a CM match and the other one as a 100% match. (Sometimes these are even shown as two 99% matches!) I always check that both of them have the same date and time signature. I can attach some examples for you. I always delete one of the matches to get only one 100% match that I apply to the segment. It's laborious, though, as sometimes there are lots of these double matches and I have to be deleting the doubles one by one. Can these doubles be caused by a slow server or maybe by the wrong import settings???