Machine Translation

Apostroph Group will, on request, deploy machine translation for your translations

Machines such as Google, DeepL, Systran, Lilt, Kantan, etc.
Everyone has at some point used Google Translate, DeepL or one of the other leading machine translation tools to quickly find out what lies behind those interesting-looking Chinese or Russian characters, or to understand the contents of a foreign-language email. And “quick” is what machine translation is all about – that, and the fact that the service is often free.

Machine-based tools will continue to evolve – with major implications for the day-to-day work of professional translators. That said, it’s not necessarily a question of either/or: the combination of human and machine is often the best way forward.
Generally speaking, you should ask yourself what you need the translation for. Do you want to know what a certain text says? Do you need to understand the nitty-gritty of a contract? Do you want to send a friendly message to your foreign work colleagues? Are you seeking to market your new product? Would you like to send some localised correspondence?
Not every text benefits from machine translation. We can talk you through your requirements and point up the pros and cons of the various options.

Computer-aided translation (CAT) applications began emerging in the 1980s. As a result of the technology, the translation sector witnessed a real revolution in the 1990s: computers could now remember “translation units” (source language segments paired with their translations) generated in earlier jobs, and words and phrases from a glossary database could now be recognised in the source language and their translations offered. The effect was profound: after all, human translators are incapable of remembering every detail of their past work. It meant that consistency in translation was now deliverable without hours of browsing through past work. And there was more: clients were now no longer dependent on a single translator amassing relevant knowledge about them over the years. Instead, because CAT allows translators to share their translation units with others, individual translators could take a break or retire without this affecting overall service delivery.

Just as CAT revolutionised the market, neural machine translation based on artificial intelligence is now the new kid on the block. On request, Apostroph Group can combine your valuable data with machine translation. This involves creating the client project in the CAT application as usual and populating it with any pre-existing translation units from the client’s translation memory. Then the machine translation aspect comes into play: where no translation exists in the memory, the machine translation database is accessed and the segment translated accordingly. 

Once the project has been populated with the additional machine translation content, your text is forwarded to a native-speaking proofreader for post-editing. This involves manually – “humanly” even! – checking your text and making amendments to the machine-translated units as necessary. We comply with the ISO 18587 standard for post-editing services; in fact, we also add a standard of our own: for its proofreading requirements, Apostroph has, for the past 25 years, made exclusive use of specialists working in their own native language.

We can advise you on machine translation – personally and without a machine!
Apostroph Group: taking you forward into the future! After all, new technology does not preclude premium quality. Aside from faster delivery times, our clients can also benefit from even better value for money. Why not call today and ask for a bespoke quote? Contact us for further information.