It can be very difficult to see the difference between all current technologies, particularly if you are unfamiliar with the world of translation. What does the future hold for professional translators in a world where the current buzzwords are “big data” and “artificial intelligence”? In this article, we’re going to demystify the concept of automatic translation tools.
The advantages of automatic translation
Automatic translation technologies have greatly improved, particularly thanks to deep learning (an automatic learning method based on “neurones”). Automatic translations are no longer based solely on statistics (probability of translation of a word via the analysis of thousands of texts). We’re talking about the most well-known, Google Translate and Reverso, that now make it easy to understand a short text in a foreign language with just a few clicks. They can be very helpful, because not only are they free and instant, they also provide a base for understanding the text.
You may have noticed that automatic translation is increasingly integrated in the tools we use every day, whether on Facebook, Google Chrome, TripAdvisor, Skype or those aimed at businesses.
Immediacy is its greatest advantage. Let’s take the example of software used by customer services: the integration of Google Translate enables advisors to instantly understand messages sent to them so the customer can be redirected to the right person, e.g. the customer may be asking for a refund, requesting information about a product or about the loyalty programme. This ensures that the processing time and quality of service are similar, irrespective of the language used by the customer.
Automatic translation is also perfect when travelling abroad, whether to manage the practical aspects of booking accommodation, reading road signs or understanding a restaurant menu. We suggest that you try the Android app launched two years ago that provides a visual translation in real time:
If you want to get an overall understanding and a general idea of the context and meaning of a phrase or short, fairly simple text, automatic translation is ideally suited. However, beyond that, in spite of technological progress and the efforts of researchers and developers, it remains severely limited when applied in the world of business.
The pitfalls of automatic translation
We can’t say it often enough, accuracy and quality are key requirements in a translation. There are three essential elements to a translation: the context and the meaning of the paragraph, plus the style.
Let me tell you a little story, a few months ago, an American friend of mine posted a photo on Facebook of her baby with a huge smile, with the caption “Charlie, cheese!”. Facebook’s translation: “Charlie, du fromage!” in French… Lost in translation!
Fortunately, that was in a private context where it didn’t matter that much. However, in a business context, this type of misunderstanding can have serious consequences for the reputation of your brand with a knock-on effect on turnover. The English word “mug” has several meanings and can be translated into French as “tasse”, “agresser” and also “dupe”! Just imagine the fallout there could be if an online trader were to rely solely on an automatic translation engine. As all English speakers are aware, one word can have several meanings.
A 2016 Shotfarm enquiry showed that 40% of online shoppers abandon their baskets because of poor product descriptions1. In addition, ¼ of them returned products received because they failed to match the information in the product description. The quality of the product description and accuracy of translations is therefore vital.
A few years ago, translation engines only translated a series of individual words, now they can translate whole phrases. But what about two sentences in a row or a whole paragraph? As the director of the CNRS, François Yvon explained recently, “a whole heap of linguistic phenomena can extend over several sentences. […] when working from English to French, a machine doesn’t know whether to translate “it” as “il” or “elle” because the relevant information may be found three or four sentences before. It’s difficult, because information needs to be memorised over a long distance.”
In a text, sentences aren’t independent from each other, the style and information is dispersed throughout. Therefore, unless this kind of information is memorised, it’s impossible to produce an accurate automatic translation.
Technology at the service of human translators
When you know that a native-speaking professional can translate around 2,000 words a day, automatic translation is undoubtedly the winner in terms of response time. But as you will have realised, it can be very risky to entrust a business translation solely to a robot: the costs of a poor translation, although sometimes unseen, are nevertheless very real. More returned products (for online traders), bad reputation, lack of professionalism, customer dissatisfaction…need I go on?
Human translation is certainly more expensive than automatic translation, but the quality is infinitely superior. In addition, the impact of a good translation is considerable: a higher conversion rate, increased customer loyalty and better natural SEO on the different marketplaces…
That said, we don’t think there’s a war between machine and human translation. They are quite simply complementary: automatic translation drastically reduces the time required for a translation and so optimises costs. it can provide a base for professional translators who can then rework the text, reformulate, stylise, and above all adapt it to the context and consumers in the target countries.
1 http://www.eweek.com/small-business/online-shoppers-blanch-at-inaccurate-product-descriptions
With Textmaster, you have access to the power of a network of verified, expert translators and the best translation tools to guarantee you maximum responsiveness and the best rates whilst preserving the highest level of quality. Request a translation quote by clicking here.
