Translation software is designed for translators and translation managers. These programs can include a wide array of features, including translation memory, glossary handling, file and project management, file conversion, and file alignment, to create translation memories. Besides word processors, these are probably the tools most widely used by translators.
In simple terms, a translation tools processes the source language files to create the standard format that the software uses to handle the text. This converts the files from their native format into the coded format called xliff that is used to translate. In this format, the original text is segmented into translation units (or TUs, which are most often sentences), and the format and structure information is coded so that it can be reproduced in the final translated version of the document. One of the great strengths of these tools is their ability to bridge across different formats, from MS Word and html, to PowerPoint, Excel, PDF, and other, more difficult types of files.
Translation software also creates and maintains translation memories (TM), databases containing the source and target language translation units for previously translated content. As the translator works through the TUs that make up the file, usually in a specialized editor that queries the translation memory to produce matches for the translation unit currently being worked on, more TUs are added to the TM.
As the translator works, he or she is presented with matches, which can be full (an identical sentence, or a 100% match) or partial, also called a fuzzy match (e.g.: "the white cat" might give a 90% match for "the black cat"). These tools can speed up the translator's work, especially when the source text contains many repetitions. There are also facilities for aligning existing pairs of source and translated files to build content for translation memories. This can really be useful to update previously translated material, as anything that was not altered will result in a high fuzzy match, often 100%.
When the translatable content includes few repetitions, or the TMs don't have much relevant content, the benefits of can be marginal. When deciding whether to use translation software or some other tool, you should examine the material at hand. Often, specialized texts such as instructional materials are unlikely to produce more than a handful of good matches, with the exception of any repeated instructions and interface elements. The more unique and specialized the material, the less likely it is for the TM to be beneficial.
Programs such as SDL Trados Studio and MemoQ can be costly, on the order of several hundred dollars, and involve a fairly steep learning curve. There are also many open source and free options available, such as OmegaT.
Contracting with translators or agencies that use TM tools is likely to produce some benefit, since it is common practice for full matches and high fuzzy matches to be discounted. However, due to their cost and specialized nature, asking non-professionals such as volunteers, scientists, or content specialists to acquire or learn to use translation software is probably not a viable option.
These are some of the key advantages of translation software tools:
• Leveraging previously translated content
• Many file conversion options
• File alignment
• Translation features, such as glossary handling, concordance searches, etc.
• Project management options
They can also have some disadvantages:
• They can be costly.
• They can be difficult to work with, and require at least some level of training.
• They can be distracting, and users may feel constrained by them, at least at first.
• They use specialized terminology that may not be clear to the lay user, and often similar features are named differently from what is used in other tools.
This wikipedia entry contains information on the xliff format, and links to many translation tools: https://en.wikipedia.org/wiki/XLIFF