memoQ - Low match percentage in identical segments with changing figures - And how to solve it! ناشر الموضوع: Tomás Cano Binder, BA, CT
|
|---|
Dear colleagues and nevertheless friends,
Kilgray's support people just solved a problem that was really puzzling and was reducing our productivity with memoQ in segments containing numbers. The problem happened when a new segment was identical to one present in the TM, except for figures with a different number of digits.
You know that if you have the translation of this segment in the TM...
Set of 5 screwdrivers with magnetic tip
...and you t... See more Dear colleagues and nevertheless friends,
Kilgray's support people just solved a problem that was really puzzling and was reducing our productivity with memoQ in segments containing numbers. The problem happened when a new segment was identical to one present in the TM, except for figures with a different number of digits.
You know that if you have the translation of this segment in the TM...
Set of 5 screwdrivers with magnetic tip
...and you translate a new segment where only the figure changes:
Set of 9 screwdrivers with magnetic tip
memoQ gives you a 99% match, and even adapts the number automatically. In theory, this should be also the case if the figure has a different number of digits, like here:
Set of 12 screwdrivers with magnetic tip
For these segments with more or less digits in the figures, we got as low as 70%, and never a 99%. As it happens, Kilgray discovered today that, if your memories are rather old, some internal issue with the current versions of memoQ causes digits to have an incorrect weight when matching.
The solution to this problem is then, basically, to simply use the Repair function with the TM. 99%'s are back!
I just thought I'd post this, as other users of memoQ with older TMs could be having the same issue.
Good luck! ▲ Collapse
| | | | Stepan Konev 
الاتحاد الروسي
Local time: 08:56
أنجليزي إلى روسي
| | | | Productivity increase or cosmetic improvement? | Sep 16, 2016 |
Nice of you to let us know about this issue and its solution.
I'm still on MemoQ 2013 R2, but I'll check next time.
For the sake of arguing, I don't get how productivity was affected:
If anything, the translator only gained from the issue, by translating each incorrect match more quickly than anticipated by the match rate. Maybe they could even make a killing, for instance if every other sentence was different and the exact replica of the previous one except for the num... See more Nice of you to let us know about this issue and its solution.
I'm still on MemoQ 2013 R2, but I'll check next time.
For the sake of arguing, I don't get how productivity was affected:
If anything, the translator only gained from the issue, by translating each incorrect match more quickly than anticipated by the match rate. Maybe they could even make a killing, for instance if every other sentence was different and the exact replica of the previous one except for the number of digits, like
2 birds in a cage
20 birds in a cage
He took 3 pills yesterday
He took 32 pills yesterday
The building is 2 years old
The building is 54 years old
Etc.
Or was productivity reduced overall in catalogues because of less relevant candidates higher up in the matching list?
Thank you,
Philippe ▲ Collapse
| | |
|
|
|
Philippe Etienne wrote:
Or was productivity reduced overall in catalogues because of less relevant candidates higher up in the matching list?
Exactly. I should have mentioned it explicitely. Because of the match of the good segment going down to 70%, memoQ would propose a number of other segments with higher matches (in the range of 80% upwards) but which required changes in words, i.e. actual editing.
Repeat this situation a hundred times in a large job, and you will have easily lost half an hour of your working day.
| | | |
Thank you for your reply, I sympathise. Knowing that you could be faster but can't because of a glitch is annoying.
Had I known this issue when working under pressure, I may even have tried to go back to Trados.
Now if you charge by the weighted word (linked to a sensible discount grid), you're paid according to the match the machine gives you, so there's not much productivity loss there in terms of earnings per hour.
Philippe
| | | |
Philippe Etienne wrote:
Now if you charge by the weighted word (linked to a sensible discount grid), you're paid according to the match the machine gives you, so there's not much productivity loss there in terms of earnings per hour.
Yes, indeed. However, very many of my customers do their analyses themselves, many in Studio, and pay according to that. While in Studio these 99% matches would probably be counted as such, in memoQ they caused the work of a 70% match, i.e. there was a bit of unpaid extra work.
I am aware that we are discussing really minimal amounts and a marginal number of segments, but nevertheless... I thought that was memoQ's normal behaviour, and lived with it for years. Initially, I had asked Kilgray about any setting I could adjust to give less weight to the number of digits in figures, and it was quite a shock to learn that it was incorrect behaviour and that it could be fixed.
| | | | To report site rules violations or get help, contact a site moderator: You can also contact site staff by submitting a support request » memoQ - Low match percentage in identical segments with changing figures - And how to solve it! | Wordfast Pro | Translation Memory Software for Any Platform
Exclusive discount for ProZ.com users!
Save over 13% when purchasing Wordfast Pro through ProZ.com. Wordfast is the world's #1 provider of platform-independent Translation Memory software. Consistently ranked the most user-friendly and highest value
Buy now! » |
| | Trados Studio 2022 Freelance | The leading translation software used by over 270,000 translators.
Designed with your feedback in mind, Trados Studio 2022 delivers an unrivalled, powerful desktop
and cloud solution, empowering you to work in the most efficient and cost-effective way.
More info » |
|
| | | | X Sign in to your ProZ.com account... | | | | | |
Login to reply/comment 
