Skrivanek Blog

No form of communication is more natural than watching and listening to another human being explain something. These days, millions of people seek videos online to display products they are might purchase or to offer support for those they own. These videos will most likely include your company’s products whether or not your company created them.

The international commerce conversation is shifting from a focus on text generated by company marketing specialists, to one with more interactive elements at its core: direct customer feedback about what is and is not desirable, unadorned facts shared on social media and customer-generated videos about what your product is, and your own professional multimedia pieces.

Translation industry analysts expect audiovisual translation (AVT) to grow at four times the rate of other industry segments in the coming years. AVT includes subtitling, dubbing, voiceover, closed captioning, apps and video games. Almost 5 billion videos are watched on YouTube daily, and 80% of those are outside the U.S. The market trend is clear. In addition to customer desires for audiovisual communication, some governments have increased regulatory requirements for closed captioning and subtitling.

The challenge to company leadership is to think of the core message for their marketing as the pure idea – not the verbal texts. Translating that core idea directly into multimedia forms (including but not limited to text) speeds up and streamlines the process of reaching your global customers. In the words of Jim Compton: “text and language are just characteristics, or dimensions of content’s form.”*

This paradigm shift involves thinking about your product and the marketing and sale of it in global terms from the very beginning. Traditionally, companies have thought of their own nation’s population as their ‘first’ market, with ‘foreign’ markets as potential customers down the line. But today so many networks of commerce aren’t limited by the same, old geopolitical or even linguistic boundaries. The question becomes how to describe your product to anyone in the world in a form more pure and immediate than written words.

Improved audio visual technology continues to make this easier and more cost-effective. There are obvious obstacles in localizing an online video tutorial for use of your product, for instance. As an example, it takes 30 minutes to engineer one minute of audiovisual from one language to another. Voice actors are also expensive. But already software exists that can use twenty minutes of a recorded voice, break it down into the smallest units of sound, and reconstruct new sentences from those – sentences that include words the original voice never spoke.*

The exciting options available in the realm of audiovisual localization are ever increasing. For the expertise you need in order to localize using AV concepts and technology, look to Skrivanek. Our linguistic experts will help you translate your core messages into every language, format, platform, and medium you require.

*Multilingual, Sound and Vision, Jim Compton, September 2017

**Project VoCo – Multilingual, AudioVisual Localization, Kamil Juljanski, September 2017.

 Contact us!

www.skrivanek.com

Machine Translation (MT) has been with us for several decades. The first effective form developed was Rule-Based MT (RBMT), initiated in the 1950’s. RBMT became obsolete when Statistical MT (SMT) was refined in the 1990s, and one form of SMT – Phrase-Based MT – still defines major online translation services.

Since 2014, Neural Machine Translation (NMT) has moved into the language services arena, opening the door for a potential paradigm shift. This is because the way NMT operates is fundamentally different, so the form it will take as it is used and evolves is not easy to predict. It’s described as “mysterious” in a Systran blog that attempts to breakdown in detail how it works,* and part of the reason it is so complex and difficult to explain is because the NMT seeks patterns on its own without being told exactly what to look for, and in the layers of processing its hard to detect how it makes its decisions.

SMT basically works by comparing source text ‘n-grams’ – 6-word groupings of words – to target language match possibilities. NMT, on the other hand, builds its data and methods through ‘deep learning’ processes which, as NMT’s name indicates, somewhat resemble the biological neural networks of animal brains. Rather than following task-specific programming, NMT systems approach problems by seeking connections from examples.

NMT systems run on Graphical Processing Units (GPUs), which are powerful and require a fraction of the memory that the Central Processing Units (CPUs) that SMT need. However, the training that the systems require is “computationally expensive,” Google says.**

Other drawbacks are that NMT doesn’t handle rare words well and this has hindered its efficiency. But with isolated, simple sentences, Google’s NMT “reduces translation errors by an average of 60% compared to Google’s phrase-based production system,” according to a Google abstract.**

Four NMT systems are currently available: Google translate, Microsoft Translator, Systran Pure Neural Machine Translation, and an open source NMT called OpenNMT from the Harvard NLP group. As the more sophisticated MT systems that Language Service Providers utilize incorporate NMT, you can be sure that Skrivanek will keep you informed of any new capabilities the technology might make available to you.

*blog.systransoft.com: How does Neural Machine Translation work? October 17, 2016

**2016 Google abstract

 Contact us!

www.skrivanek.com