Most professional translation agencies and freelance translators are keeping a close eye on the trends in machine translation. Sooner or later, it is generally realized, machine translation technology will have advanced to such a point that it might threaten the livelihoods of human translators, in the same way that AI and robotics are beginning to threaten the jobs of many other workers.
Up to now, machine translation has been seen as a tool to be used by professional translators to speed up their translation tasks, and has not been seen as a direct substitute for the accuracy of human translation.
The latest trends in machine translation include a technology called neural machine translation (NMT). The rapid development of NMT could potentially have a massive effect on the translation industry. NMT’s key point of difference is that it focuses on the translation of whole sentences rather than combining the translation of individual words.
How does NMT Work?
NMT works by using neural network-based tools to create more precise translation. NMT in fact uses something more akin to the way that human translators work by the use of a large, albeit artificial, neural network. NMT is able to create a single integrated model through the use of whole sentences combined together.
NMT takes the sentences in a chunk of text and converts them into language vectors in what is called the ‘encoding’ phase. The vectors are then automatically decoded into the target language. Unlike current MT technology NMT is able to deal with context based vocabulary and syntax, making it a more reliable method of translating large volumes of text.
NMT appears to be far better at translating more complex languages such as Arabic, Japanese, Chinese and Korean, all key languages for translation work.
Advantages of NMT in the Future for the Translation Industry
There are other advantages of NMT which make it stand out as a technology to watch in the near future. Trends in machine translation associated with NMT that give it an edge include its improved ability to deal with the nuances and subtleties inherent in every language. That includes gender based language, inflection and the use of formal and informal speech.
The immediate beneficiaries of NMT are likely to include:
- eLearning courses;
- Multilingual video conferencing;
- The travel industry;
- Technological support to customers;
- Fast turn-around time for legal translation e.g. for litigation purposes.