• State Suggested Answer
  • Replies 6 replies
  • Answers 1 answer
  • Subscribers 435 subscribers
  • Views 5299 views
  • Users 0 members are here
Options
Related

Statistics/Analysis - difference between Trados and MemoQ

Anna Herboly

Dear SDL Community,

I'm facing an issue with statistics / analysis.

My client uses MemoQ, I'm a Trados 2007 user.

For new projects, they send me the files and their statistics which is different from what Trados returns upon performing an analysis on the source files.

The client of course deems their statistics to be correct.

Do you possibly have an idea where this discrepancy is rooted and which settings I should change to obtain the same figures as my client?

 

Here's the client's analysis:

Screenshot of Trados Studio statistics showing categories like Duplicates, Translated, Exact Match, No Match, and their respective segment and word counts.

 

And here's what Trados returns - for the same package of files:

Screenshot of an analysis report in SDL Trados Studio with summary, settings, and detailed overview of match types including percentages and word counts.

As you will see, not even the number of segments is the same in both analyses.

However, the client would like me to submit my quote based on the analysis.

Any hints on this please-? Your help is much appreciated - thanks very much, best,

 

Anna

Generated Image Alt-Text
[edited by: Trados AI at 2:14 PM (GMT 0) on 28 Feb 2024]
emoji
Rate translation
×
Suggest better translation
×
Moderator UI
×
  • Get AI Suggestion
Parents
  • 0
    Still, there is something wrong in your post... you say that you are using Trados 2007, but screenshot is from Trados Studio...
    Plus, if you get MemoQ package, shouldn't it be already segmented and converted to MQXLIFF? If yes, and if you are indeed using Studio, why don't you simply open the MQXLIFF? Then you should not have differences in number of segments, as the MQXLIFF should be already segmented.
  • 0 in reply to Evzen Polenka

    Any mqxliff file inside a Studio project gets segmented again (meaning that sentences could get split in sub-sentences), unless you use some empiric workaround.

    Also Memoq out of the box doesn't count tags and (as far as I know but would be better to check) considers a word anything between two spaces, thus producing a different word count from Studio.

    Also when you Analyze it against the same TM (via tmx import) you get different results, usually in favor of memoq since you started the project with it.

    Besides in the screenshot I see two "Exact Match" rows and not a "Fuzzy" row, so I guess there's something wrong with it.

    I know anything of this is a solution, but IMHO looking for a better arrangement with the agency could be more productive than trying to get the same analysis with two completely different tools.
Reply
  • 0 in reply to Evzen Polenka

    Any mqxliff file inside a Studio project gets segmented again (meaning that sentences could get split in sub-sentences), unless you use some empiric workaround.

    Also Memoq out of the box doesn't count tags and (as far as I know but would be better to check) considers a word anything between two spaces, thus producing a different word count from Studio.

    Also when you Analyze it against the same TM (via tmx import) you get different results, usually in favor of memoq since you started the project with it.

    Besides in the screenshot I see two "Exact Match" rows and not a "Fuzzy" row, so I guess there's something wrong with it.

    I know anything of this is a solution, but IMHO looking for a better arrangement with the agency could be more productive than trying to get the same analysis with two completely different tools.
Children