The time for laughing at the clumsy machine translation is over. The development and application of artificial intelligence (AI) to machine translation is changing everything.
Advances in machine translation (MT)
In the era of digital technology and artificial intelligence, symbolic and statistical systems, which are based respectively on linguistic rules and corpora of translated texts, have become almost obsolete.
The emergence of various machine translation tools, based on artificial intelligence and neural networks, has marked a key step in the evolution of machine translation tools. For example, voice machine translation and adaptive machine translation are likely to occupy a prominent position in the translation landscape in the years to come.
Voice machine translation
Real-time voice translation , or speech-to-speech translation (S2S), makes use of the latest advances in artificial intelligence, such as deep neural networks.
It is necessary to go through three distinct stages to translate an oral speech from source language to target language:
- Voice recognition transcribes oral speech, which can then be processed by machine translation systems. Voice recognition is only possible thanks to deep neural networks, which work similarly to the human brain.
- Machine translation translates the transcription into the target language.
- The oral message is synthesized, transforming the written translation into speech.
Skype has entered the voice machine translation market: Skype Translator translates voice conversations in real time. This service is available in seven different languages.
Adaptive machine translation
AdaptiveMT , was developed by SDL , and draws on both machine learning and translation.
In concrete terms, how does it work?
AdaptiveMT can rightly be called a private machine translation engine. That is, the translator has his/her own machine translation engine at his/her disposal. The self-learning machine translation engine adapts and learns in real time as segments are translated using the software.
All changes are therefore made instantly in the text which makes the text more coherent and adapted to customised analysis. Data is the key to this system, as it drives the analysis. In short, if the subject matter to be analysed increases, translations will approximate human translations in terms of quality and fluidity.
What is the future of machine translation?
Today, AI development has a bright future, but machines are not yet ready to replace professional translators.
With the abundance of multilingual information, translators are finding it increasingly difficult to meet the exponential demand for translation services. By using the tools mentioned above, professionals in the field remain in control of the creative process, while increasing their productivity.
These advances can allow translators to focus on texts to increase added value. However, it is undeniable that the boom in machine translation, which is becoming more and more reliable and precise, goes hand in hand with evolution in the field of post-editing.
Future advances in machine learning will allow for profound changes in communication and translation techniques on a global level.
Translated by Willem Beckmann