Forum: CAT Tools Technical Help
Topic: Best CAT Tool to Utilise your past translations
Poster: DZiW
Post title: no fat lucky cat got the tongue
A colleague of mine decided to go with the time and modernize the work-flow via so fabulous Computer Aided/Assisted Translation, because they told her it would be possible to re-align and re-use even her 20+ year-old translations in any pairs...
Some language pairs are rather specific, and what is common in one is not ok for the other: wording, tempo, strings/paragraphs length/patterns, let alone a client's instructions and the target audience. Shortly, it appeared almost impossible to align easily and properly, say, 1200+ English segments with some 900 Russian varied sentence type segments or even 400 Japan segments with 300 French ones. She conveyed the idea, yet no sentence for sentence equ, merging and moving the meaning to the bare bones.
When I explained it to her, the alignment relies on the similarities of parallel structures, she exclaimed that it's absurd for no translation could and should be identically segmented, and inquired about another 'better' tool. I must admit, after the final edit my translations also may differ, but no more than a few sentences (if the client permits to combine original short/orphan or delete not so relevant ones). However, she did like the centralized database/backup concept.
Meanwhile I had to refer to the paragraph-based segmentation instead of sentence-based, which significantly lowers the fuzzy, but still works... occasionally.
Topic: Best CAT Tool to Utilise your past translations
Poster: DZiW
Post title: no fat lucky cat got the tongue
A colleague of mine decided to go with the time and modernize the work-flow via so fabulous Computer Aided/Assisted Translation, because they told her it would be possible to re-align and re-use even her 20+ year-old translations in any pairs...
Some language pairs are rather specific, and what is common in one is not ok for the other: wording, tempo, strings/paragraphs length/patterns, let alone a client's instructions and the target audience. Shortly, it appeared almost impossible to align easily and properly, say, 1200+ English segments with some 900 Russian varied sentence type segments or even 400 Japan segments with 300 French ones. She conveyed the idea, yet no sentence for sentence equ, merging and moving the meaning to the bare bones.
When I explained it to her, the alignment relies on the similarities of parallel structures, she exclaimed that it's absurd for no translation could and should be identically segmented, and inquired about another 'better' tool. I must admit, after the final edit my translations also may differ, but no more than a few sentences (if the client permits to combine original short/orphan or delete not so relevant ones). However, she did like the centralized database/backup concept.
Meanwhile I had to refer to the paragraph-based segmentation instead of sentence-based, which significantly lowers the fuzzy, but still works... occasionally.