Translation memories: OmegaT

OmegaT is an open source translation memory that register equivalences between two languages. Like the translation memories in general, with OmegaT the translators are able to keep a lot of documents that will be very useful in future translations. In the program there are three windows: the first window is where you will see the source text, in the second one the program will show you the possible equivalences and below this one it will be a glossary.

To start to work with OmegaT you have to click on “Projects” and then create a “new project”. You will have to keep a new folder where your work will be kept. In the folder you will have the documents you want to introduce in the translation memory. You can use OmegaT to start without any previous translation or maybe you want to take advantage of other translations you have already had. However, if you want to do this last, you must use before a program, like for example bitext2tmx, that will help you to align the source text and the translated text, because OmegaT doesn't have this option.

You will find bitext2tmx here (bitext2tmx.zip). Once you have installed that program and it is open, you will see the next image:

And now you must follow the next steps:

1. Click on “file” and select the source text, the source language and the target text and the target language.
2. It will appear both texts on the program and you will have to verify if all the sentences in the source text correspond with the sentences in the target text.
3. If the sentences don't correspond, you will have to modify the target text with the buttons “join”, “delete” or “split”. 

It is very important that the texts are in format TXT.

This alignment will be useful in your following translations of texts of the same or similar subject, because OmegaT will show all the coincidences and you will have to select a partial coincidence, a total coincidence or you will have to rewrite all the sentence. In any case, if you work with repetitive texts, this will speed up your work.

We return to OmegaT, and now you will have to open the project and select the language of the original texts and translated texts and all will be stores in the new project and below you have to indicate too the folder where all the files that will be generated in the project will be saved (the original text and the target text that have aligned with bitext2tmx, the translation memory that is generated and stored, the glossaries that you add optionally and the translated text).

To introduce the text you want to translate, you will have to press on "import source files" and select it.

It is important that you check, that the folders you have created in your project contain the texts and memory that was generated in bitext2tmx. 

In the main frame you will see the original text and in the right frame will appear the possible matches from the memory, with the text you want to translate and the percentage of coincidences. If you keep more texts in your memory, it will find more matches in similar texts. 

And now you can start to translate. To go faster, and once you are upon the segment that you want to translate, you can press CTRL + R and it will be replaced by the translation automatically. If there are more than one match, you can choose the most suitable pressing CTRL + the number of the coincidence. 

Don't forget to save your work and finish "creating final documents" from the menu "project".

And now you can translate with OmegaT!