Transcad provides translations of both printed and online
documentation (manuals) for various software products into Czech/English.
Original formats, text elements, and styles are preserved. Supported
formats include DOC, RTF, PDF, HTML, JS, FM, and MIF.
We use the CAT (Computer Aided Translation) technology
with TM (Translation Memory) feature. The concept of this approach is very
simple. When the text is translated, it is included in the database that
is used in next translation cycle (upgrade/update). The main volume of the
text is therefore translated just once. Identical segments are translated
automatically and only new/modified text is translated manually. As you
can see, this technology saves both time and financial resources required
for the translation to be completed. If you are interested in the concept,
see the overview below.
The TM Technology
Let us briefly outline the principle of TM database. For this
purpose, we will use a sample application called SampleSoft that needs to localized into Czech language. However, the concept remains
the same for any sort of text.
Phase 1 - SampleSoft 1.0
Localization
SampleSoft 1.0 has been launched. The very
first translation/localization of the SampleSoft 1.0 is required. In this
phase, the entire volume of text must be paid by the customer as it is
treated for new text. However, even now some costs are saved - repeated
segments are paid just once. This applies to common phrases like "Click
OK." or "Save the file and proceed to next excercise".
Billingual pairs are saved into TM1 (translation memory) database. The
format is "English segment = Czech segment". See the
chart below.
SampleSoft 1.0 Localization
text that needs to be translated
(non-repeating segments) text that gets translated from the database
(repeated segments)
Phase 2 - SampleSoft 1.1 Localization
Updated
version of the SampleSoft has been launched. The text is analyzed against
TM1 translation database from SampleSoft 1.0 localization (see above). The
analysis shows that 90 % of the volume are identical and remaining 10 %
show differencies. The matching intervals are calculated and new/modified
text is translated. Billingual pairs of SampleSoft 1.1 version are stored
into TM1.1 translation database. In this phase, 90 % of the text get
translated automatically FOR FREE and only 10 % are translated and paid by
the customer. Moreover, the translation process is much faster - only
10 % of the total volume need to be translated. See the chart below.
SampleSoft 1.1 Localization
text that needs to be translated (new and modified segments)
text that gets translated from the database (identical segments)
Phase 3 - SamleSoft 2.0 Localization
SampleSoft
2.0 has been launched. The text is analyzed against TM1.1 translation
database from version 1.1. The analysis shows that 85 % of the volume are
identical and remaining 15 % show differencies. The matching intervals are
calculated and new/modified text is translated. Billingual pairs of
SampleSoft 2.0 are stored into TM2 translation database. In this phase, 85
% of the text get translated automatically FOR FREE and only 15 % are
translated and paid by the customer. Moreover, the translation
process is much faster - only 15 % of the total volume need to be
translated. See the chart below.
SampleSoft 2.0 Localization
text that needs to be translated (new and modified segments)
text that gets translated from the database (identical segments)
text that needs to be translated
(non-repeating segments)
text that gets translated from the database
(repeated segments)
