Technical forums » MemoQ support »
Question about Repetitions
Thread poster: Donglai Lou (X)
Donglai Lou (X)
Donglai Lou (X)  Identity Verified
China
Local time: 16:30
English to Chinese
+ ...
Aug 13, 2013

Dear Colleagues,

I have a question about how MemoQ counts repetitions.

With Trados 2007, we know if there are 3 identical segments, the first one will be counted as new texts and the remaining two are repetitions. however, it seems that in MemoQ, all of the three segments are included in the repetition part. if so, this might lead to unfairness.

Considering 25% of the normal rate is applicable to repetitions, if there are only 2 identical segments, usin... See more
Dear Colleagues,

I have a question about how MemoQ counts repetitions.

With Trados 2007, we know if there are 3 identical segments, the first one will be counted as new texts and the remaining two are repetitions. however, it seems that in MemoQ, all of the three segments are included in the repetition part. if so, this might lead to unfairness.

Considering 25% of the normal rate is applicable to repetitions, if there are only 2 identical segments, using Trados, you can get a weighted count of 1.25 segment, but with memoq, the weighted count is 0.5 segment.

Can anyone kindly enlighten me on this？

besides, I am not sure that anyone has noticed that the fuzzy match rates are often inaccurate.

best regards,
DL ▲ Collapse
 
preynolds
preynolds
Local time: 10:30
clarification about repetitions in memoQ Aug 13, 2013

Dear Donglai Lou,

Thank you for your interest in memoQ. However, in memoQ repetitions do not work the way you describe. We (Kilgray) were very concerned when we saw your post and checked and double checked that this was working the way it was supposed to. That is, if there are 3 segments which are the same, the first one is counted as new words and the second and third are counted as repetitions. Our tests have shown this is exactly what happens. If anyone would like to check this t... See more
Dear Donglai Lou,

Thank you for your interest in memoQ. However, in memoQ repetitions do not work the way you describe. We (Kilgray) were very concerned when we saw your post and checked and double checked that this was working the way it was supposed to. That is, if there are 3 segments which are the same, the first one is counted as new words and the second and third are counted as repetitions. Our tests have shown this is exactly what happens. If anyone would like to check this they should create a document with one sentence repeated three times, among other text which is not repeated. They should then import this document and do a statistical analysis without adding a TM. You will see that if there are three identical segments that memoQ tells you that two segments are repetitions.
As memoQ does not count repetitions in the way you describe your maths is also wrong. If there are two segments memoQ gives the same count as Trados - 1.25.
I would need more information on the fuzzy match rate issue you mention. All tools have some differences as to how they measure fuzzy matches. For examples memoQ gives a fuzzy match of 95% if the text is exactly the same but there are formatting differences. Some other tools would classify this as a 100% match.
I hope this clarifies the issues. There is more information available on http://kilgray.com/resource-center and http://kb.kilgray.com/

Thanks,

Peter Reynolds, Kilgray ▲ Collapse
 
Donglai Lou (X)
Donglai Lou (X)  Identity Verified
China
Local time: 16:30
English to Chinese
+ ...
TOPIC STARTER
Thanks a lot for clarification Aug 13, 2013

Hi! Peter,

Thanks so much for your clarification. I was not sure about it. this is why I would like to check with you to get the authoritative answer. sorry for not figuring out how to verify it. Now it is clear.

for the fuzzy match, I will give you some examples, gradually.

Now I have three ones in my archive:

Example 1:
Integer Research obligations to Sichuan Zhongzi Exhaust Purge Co., Ltd:
Sichuan Zhongzi Exhaust Purge Co., L... See more
Hi! Peter,

Thanks so much for your clarification. I was not sure about it. this is why I would like to check with you to get the authoritative answer. sorry for not figuring out how to verify it. Now it is clear.

for the fuzzy match, I will give you some examples, gradually.

Now I have three ones in my archive:

Example 1:
Integer Research obligations to Sichuan Zhongzi Exhaust Purge Co., Ltd:
Sichuan Zhongzi Exhaust Purge Co., Ltd
match rate: 75%

Example 2:
Adertising and Branding at Conference
Adertising and Branding pre- Conference
match rate: 65%

Example 3:
Logo Displayed as "silver sponser" on the front of the conference guide
Logo Displayed as "silver sponser" on the front of the conference brochure sent to 3,000+ recipients (requires booking made well in advance).
match rate 71%

my friend, a MT developer, has given me the formula you might use to calculate the fuzzy match. I believe it is mathematically sound but fails to measure the real workload.

anyway, I still believe MMQ is the very best CAT available to translators in the market.

thank you for your excellent work.

best regards,
Donglai Lou ▲ Collapse
 
preynolds
preynolds
Local time: 10:30
Fuzzy matching in memoQ Aug 13, 2013

Dear Donglai Lou,

Many thanks about your kind words about memoQ. I am sure your friend has many insights into translation technology. However, you cannot state there are problems with our fuzzy matching algorithms based on your friend's guess as to what they are. The fuzzy match percentage will be effected by the length of segment, order of words and other changes in the segments. In my opinion the examples you give are probably too short to analyse our fuzzy matching. It is more im... See more
Dear Donglai Lou,

Many thanks about your kind words about memoQ. I am sure your friend has many insights into translation technology. However, you cannot state there are problems with our fuzzy matching algorithms based on your friend's guess as to what they are. The fuzzy match percentage will be effected by the length of segment, order of words and other changes in the segments. In my opinion the examples you give are probably too short to analyse our fuzzy matching. It is more important to look at how the tool can improve the leverage you got on a consistent basis.

Good luck,

Peter. ▲ Collapse
 


To report site rules violations or get help, contact a site moderator:

Moderator(s) of this forum
Maya Gorgoshidze[Call to this topic]
Peter Zauner[Call to this topic]
Prachya Mruetusatorn[Call to this topic]

You can also contact site staff by submitting a support request »

Question about Repetitions






CafeTran Espresso
You've never met a CAT tool this clever!

Translate faster & easier, using a sophisticated CAT tool built by a translator / developer. Accept jobs from clients who use Trados, MemoQ, Wordfast & major CAT tools. Download and start using CafeTran Espresso -- for free

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 »