Forum: CAT Tools Technical Help
Topic: CAT Tool Use as Terminology/Glossary Search & Replace Tool
Poster: Samuel Murray
Post title: Mathias
[quote]Mathias S. wrote:
The desired, basic case of application would be to use in a first step a simple search & replace on the source documents with our term database. [/quote]
This is not how CAT tools work, though. CAT tools attempt to match the source text to text in a database, so if you're going to go and replace some of the source text with target text before loading it into the CAT tool, the CAT tool will have a much harder time finding matching texts. With a CAT tool, replacing terminology in the target text is done on a segment-by-segment basis. This means that you should have a CAT tool that will show the translator which terms were found in the current segment's source text, with a feature that allows the translator to easily select a translation and drop it into the segment. Most CAT tools can do that (some more effectively than others).
I suggest you try OmegaT to begin with, just to get a feel for how a CAT tool works. OmegaT's glossary pane is not the most user-friendly, and it has the unfortunate bug/feature whereby it merges glossary entries with identical source texts into a single entry on screen, which makes it harder to match the right description to the right target text, but it'll do, mostly, and especially as a free introduction to how CAT tools work or can work. What OmegaT can't do, is pretranslate non-100% matches.
Trados' glossary display isn't very nice to work with, and it is very difficult to set up a glossary in Trados. Trados, MemoQ, Wordfast Pro 3 and Wordfast Pro 5 all suffer from the design idea that the glossary pane should be unobtrusive and only consulted in an emergency. I found MemSource's glossary pane and glossary matching to be quite something, so perhaps you should look into that.
[quote]In CAT tools I've tried (DejaVUX3, memoQ) this works mostly with pretranslate functions and copy source text. [/quote]
Yes, CAT tools start off with an empty translation memory (TM) (i.e. database of translated sentences), and you add translations as you go. If there is no matching text yet, you can type your translation or insert the source text (and with inserting the source text, some CAT tools offer the feature of automatically replacing terms from the glossary).
[quote]But there are some options to figure out, like preferring exact matches only. [/quote]
This is often achieved by either setting the match threshold to e.g. 100%, or by telling the CAT tool not to automatically insert matching segments when a match is found.
[quote]I'm interested in flexions)... [/quote]
Not all CAT tools offer that (OmegaT does, for example).
Topic: CAT Tool Use as Terminology/Glossary Search & Replace Tool
Poster: Samuel Murray
Post title: Mathias
[quote]Mathias S. wrote:
The desired, basic case of application would be to use in a first step a simple search & replace on the source documents with our term database. [/quote]
This is not how CAT tools work, though. CAT tools attempt to match the source text to text in a database, so if you're going to go and replace some of the source text with target text before loading it into the CAT tool, the CAT tool will have a much harder time finding matching texts. With a CAT tool, replacing terminology in the target text is done on a segment-by-segment basis. This means that you should have a CAT tool that will show the translator which terms were found in the current segment's source text, with a feature that allows the translator to easily select a translation and drop it into the segment. Most CAT tools can do that (some more effectively than others).
I suggest you try OmegaT to begin with, just to get a feel for how a CAT tool works. OmegaT's glossary pane is not the most user-friendly, and it has the unfortunate bug/feature whereby it merges glossary entries with identical source texts into a single entry on screen, which makes it harder to match the right description to the right target text, but it'll do, mostly, and especially as a free introduction to how CAT tools work or can work. What OmegaT can't do, is pretranslate non-100% matches.
Trados' glossary display isn't very nice to work with, and it is very difficult to set up a glossary in Trados. Trados, MemoQ, Wordfast Pro 3 and Wordfast Pro 5 all suffer from the design idea that the glossary pane should be unobtrusive and only consulted in an emergency. I found MemSource's glossary pane and glossary matching to be quite something, so perhaps you should look into that.
[quote]In CAT tools I've tried (DejaVUX3, memoQ) this works mostly with pretranslate functions and copy source text. [/quote]
Yes, CAT tools start off with an empty translation memory (TM) (i.e. database of translated sentences), and you add translations as you go. If there is no matching text yet, you can type your translation or insert the source text (and with inserting the source text, some CAT tools offer the feature of automatically replacing terms from the glossary).
[quote]But there are some options to figure out, like preferring exact matches only. [/quote]
This is often achieved by either setting the match threshold to e.g. 100%, or by telling the CAT tool not to automatically insert matching segments when a match is found.
[quote]I'm interested in flexions)... [/quote]
Not all CAT tools offer that (OmegaT does, for example).