Let’s take a quick look at the so-called computer aided or assisted translation systems (abbreviated as CAT). Their basic linguistic resource is a text database of previously completed translations. In addition to the original text and the corresponding translations, each pair of sentences is saved, along with some additional information, such as language codes, identification of the author, date and time of translation, the domain, reference to the original file, and much more. Such database entries are referred to as translation units (TU). The set of translation units is then referred to as a translation memory (TM)).
For every sentence translated, CAT tools try to find an identical or similar entry in the translation memory. The matching is calculated at a percentage, e.g. the sentences “Paul is riding a bike.” and “Peter is riding a bike.” are a 75% match. If a similar translation done previously is found in the translation memory (the tools most commonly aim at between 75% and 100% similarity), the system informs and translator and offers the translation that is stored in the translation memory as a hint for the possible translation of the new original text. It is up to the translator to then accept or edit the suggestion.
The first CAT tools were commercially used in the 1990s, but they became more generally widespread only in the first decade of this century. The main pioneers included Trados Workbench and Wordfast, which were popular among users mainly because they were user friendly and were installed as simple add-ons to Microsoft Word.
However, the advantages of working in MS Word were gradually outweighed by the logical disadvantages – for example, it was only possible to translate a relatively limited range of file formats this way – including, for example, MS Word, MS Excel, MS Powerpoint, and HTML.
Towards the end of the decade, most translation technology providers started developing their own translator editors, usually referred to as translation studios. This was the time when MemoQ, from the Hungary-based company Kilgray, was probably the most famous tool. It was a desktop application with its own translation editor that offered users some great features for the easier translation and editing of texts. In addition to MemoQ, the market leader Trados, which was taken over by SDL in 2005, was also winning a strong market share at this time.
After 2010, we saw a strong tendency among various technologies to transform from desktop applications to web applications, and the same trend was also true in the translation industry. There were three powerful tools that emerged and gained significant market share over the next few years – Memsource, Wordbee, and XTM. Both technologies have their strengths and weaknesses.
While desktop applications require installation, regular updates, and compatibility between individual versions, their advantages include the ability to work offline, greater robustness, and, for many users, a larger sense of privacy. Web applications, on the other hand, boast accessibility via standard browsers, cross-platform compatibility, and no need for manual updates. However, especially in the early stages of their development, these tools struggled with performance, stability, and uptime issues.
Today, CAT tools have become a standard part of translation processes, and most translators and translation service providers can hardly imagine their work without them. CAT tools are ideal for professional technical translations, for software localization, and also for the translations of legal texts (generally in texts with a high degree of repetitiveness). Some CAT tools also make it much easier to manage express or large-scale translation projects where teams of translators, consultants, and proofreaders need to work together and coordinate their efforts.