Comparison of CAT software

Before I try to elaborate on my experience and compare MemoQ, across and OmegaT I should say a few things on advantages and disadvantages to use CAT programs.

General remarks

It seems to be obvious that you benefit more from CAT, if the text(s) you are translating includes a high number of repetitive phrases and sentences. I can hardly imagine that it makes much sense, if you are translating literature.

One should also bare in mind that shorter sentences will lead to quicker results. This means that a person composing a text knowing that it will be translated should make short sentences, rather than long ones, which in the Turkish language for instance is often seen as an indication of high academic value.

The problem with the segments

The segments for translation usually are determined by full stops. All three programs offer the opportunity to declare additional segmentation rules. The main exceptions are usually included. For the English language you would not have the end of a segment after expressions such as Dr. or No.

MemoQ offers the possibility to join segments (if a sentence was split, for instance). In across you can declare exceptions for each segment. I would not say that it was easy, but since OmegaT had no segmentation rules for Turkish I had to create some. After a while I was able to declare exceptions a) if only the initials of names are given (e.g. H.P.) or b) if the full stop after a number means first, second etc. (common both in Turkish and German). The other programs should have the same opportunity (to declare exceptions), but I did not try to find out how to do that.

Glossaries

All three programs offer the possibility to build your own glossarie(s). Glossaries can serve more than one purpose. Besides the usual ways of looking for vocabulary that you are uncertain about (or tend to forget) you make entries to the glossaries hoping that the term will show up when it appears in a segment. If you enter these terms into your glossary you will be reminded of possible translation(s) of such a term. Should it be a short word the possibility to insert it into your text (by shortkey or a click) is not really necessary.

The second possibility is that certain phrases often appear in the texts for translation. An example from the texts of the Human Rights Foundation of Turkey would be "propaganda for an illegal organization". Whether you are translating from Turkish to English (the phrase in Turkish would be needed) or English to German (you have to translate the phrase once) it would be nice to have the opportunity to include the complete phrase into the translation.

Glossaries in OmegaT

It looks a bit complicated to build a glossary in OmegaT, but it should not be a real problem. Unless you try a script of Samual Murray in order to have a shortkey for new entries to your glossary you manually have to create one (or more) glossaries. You can do this in a simple text file by using tabs. First you enter the text in the source file, hit TAB ->| and enter the translation (target file). With another TAB ->| you can add further information.

It may be unusual, but you have to save the text-file with the extension *.utf8. All of this is explained on the page concerning OmegaT.

There are intentions to create the possibility to include matches in the glossary into the translation, but for the time being you can only "copy and paste" the hits into your text.

Glossaries in MemoQ

I personally found the creation and use of glossaries in MemoQ most convenient. First you have to create a glossary. This is done in the project manager under term databases. You can use any name you like. Afterwards you have to assign the dictionary to your project or rather the glossary that you want to edit has to be assigned to the project you are working with.

During translation you can mark any word, expression or phrase either in the source or in the target text (or in both) and once you press CTRL+E a window open and the text that you highlighted will appear (according to what you marked either in the source text, target text or in both sections). There are further possibilities that you can use (add an example, specify grammatical information etc.) but there is no obligation to do so. Just press OK (or hit Enter) and you are done.

As soon as the term that you entered appears in the source text its translation will be offered to you for inserting. If there are more than one expression in the source and/or part of the text seems to be a fuzzy match then the additional proposals for inclusion will also be listed in numbers. You can, of course copy and paste the phrases and terms into your translation, but it is much easier to press CTRL+the number of the offered item.

Important: Any match found in the source text must appear as translation in the target text. Otherwise you get a warning. For details see the page on MemoQ

Glossaries in across

The glossaries in across are called crossTerm. You can open a glossary with the corresponding button or by press Alt+1. Once you have highlighted a term you can insert it either using the right mouse and the corresponding option or by press CTRl+Alt+1. You can specify further options. Make sure that you have specified the field "Usage". Before you save the new entry you should also mark "released" so that the term gets ready for inclusion. You have to do this for both languages that you are using.

The rest is pretty much similar to what was said on MemoQ. You can insert a term by pressing CTRL+number of item or with a double click of the mouse.

MemoQ, across and OmegaT

Comparison of the programs

The programs tested were:

Program	Source files	Specifics	Price (full version)
OmegaT	txt, html, OpenOffice	pretty fast, glossary of no great use	GPL (free)
MemoQ	MSOffice etc.	spell check may be slow, glossary easy	620,00 €
across	almost anything apart from PDF	rather complicated	970,00 €

Both, OmegaT and MemoQ can be called intuitive. Though it is useful to read some instructions, it is also possible to use these programs from scratch (as is) and slowly find out the possibilities later. MemoQ seems to be more sophisticated, but the beginners may find it easier to handle OmegaT. As far as the reports of the HRFT are concerned after just having translated the reports of one week there were already several fuzzy matches.

While the handling of tags in MemoQ may be annoying, the impossibility to include matches from the glossary is something missing in OmegaT.

During the testing phase all features of MemoQ were available. That may change after 45 days. In the "Help" of the program I am told that

In the Freeware Edition, you can do the following:

Create projects with one document and one translation memory. With every new project, an empty translation memory is automatically created. You cannot create, delete, import, export or edit translation memories. You can create and edit term bases. However, you cannot import or export them.

These limitations would make MemoQ rather useless. One might opt for a postcard edition which means you have more possibilities by just sending one postcard to Hungary. Alternatively, I could try to get more acquainted to across, even though it is somehow complicated to build the glossaries.