As of SDL Trados Studio 2014 SP2 the word count and search logic for words
containing certain types of apostrophes and hyphens was improved.
Studio 2014 SP2 uses an improved algorithm for processing words that contain certain types of hyphens (-
) or apostrophes
(‘
). This improvement translates into:
Lower word count
Studio no longer treats apostrophes and dashes as word separators, but as punctuation marks that link words together. This means that Studio counts elements like mom’s
, it’s
or splash-proof
, NAME_4001C
as one
single word.
Hyphens that follow the new logic:
-
Non-breaking hyphen (
‑
) -
Hyphen-minus (
-
) -
Soft hyphen (
) -
Hyphen (
‐
) -
Small hyphen minus (
﹣
) -
Full width hyphen-minus (
-
)
Apostrophes that follow the new logic:
-
Apostrophe single quotation mark (
'
) -
Modifier Letter apostrophe (
ʼ
) -
Armenian apostrophe (
՚
) -
Right single quotation mark (
’
) -
Fullwidth apostrophe (
'
)
Apostrophes that do NOT follow the new logic:
- Apostrophes followed or preceded by space. For example,
the ‘90s
orgirls’ night
both contain two words.
Dashes that do not follow the new logic:
-
Figure dash (
‒
) -
En dash (
–
) -
Em dash (
—
) -
Horizontal bar (
―
) -
Small Em dash (
-
)
Higher fuzzy matches
When searching for matches in the selected translation memories, Studio considers certain types of apostrophes and hyphens as separate delimiters inside the words they link together. This means that when comparing words where the only difference is the type of hyphen or apostrophe used, Studio only penalizes the difference in punctuation and not the entire word. This leads to better leverage and returns higher matches from the translation memory.
Example:
Let's assume the following translation unit exists in your translation memory:
English source: This is a water-proof watch.
/ German target: Dies ist eine wasserdichte Uhr.
Now let's assume you look up the following new sentence in your translation memory: This is a splash-proof
watch.
In this example, Studio SP2 will correctly identify the difference as being splash only. In previous versions (notably Studio 2011 and earlier versions), the entire word splash-proof
would have been considered to be the difference, lowering the fuzzy match value accordingly.
Important:
Re-index your existing translation memories before using them in Studio 2014 SP2. This synchronizes the translation memories with the new logic for counting and matching words that contain apostrophes and dashes.
More information on re-indexing translation memories can be found in the Online Help: http://producthelp.sdl.com/SDL_Trados_Studio_2014/client_en/SDL_Trados_Studio_Help.htm.
Information can be found here: SDL Knowledgbase
Write a comment