Top Machine Translation Stories of 2018

Top Machine Translation Stories of 2018


What a memorable year 2018 was for language professionals! It definitely did not disappoint, bringing changes the impact of which we are likely to see for many years to come. The language industry briskly moved forward in terms of technology adoption, market size and service diversification. Let’s take a look back at some of the machine translation highlights that 2018 brought us.

Publications

Not without a reason, machine translation proved to be a widely discussed topic in 2018, with several publications on the matter released throughout the year.

Slator, a news website dedicated to the global language industry, published its independent Neural Machine Translation Report, where it touched on key aspects of this technology such as estimating the expected quality of the machine translation as well as opportunities and challenges that it presents.

Woman holding a pen over a financial report

The main aim of the white paper was to share a simple and clear overview of the ways in which businesses can benefit from this emerging technology.

Intento is a startup company that opened its doors in 2018, offering its users the possibility to connect to a large number of MT engines and evaluate the different outputs. It also published a helpful comparison of different MT providers in their “State of the Machine Translation” SlideShare. The publication not only helps to understand which providers allow for best results but also the relationship between the performance and the price you pay for the service.

Providers

On the topic of MT providers, Google Translate and Microsoft Translator have been commercially available for a while but April 2018 added a new entrant to the pool of over 20 neural machine translation technology providers available at that point in time.

Amazon Translate came out of its preview mode to become a fully fledged commercial product. Although it started off with a rather limited scope of languages, it offers a scalable pricing plan based on usage and gives access to useful features such as sentiment analysis.

Jaime Punishill, Chief Marketing Officer of Lionbridge, Amazon’s commercial partner said, “We would be foolish to not imagine Amazon’s going to be a big player here.”

Technology

Another tech giant, eBay, was not so lucky with its application of machine translation. It partnered up with Webinterpret, an international ecommerce localisation company, in a programme allowing sellers to list their items on several international eBay websites.

eBay listing shown on a tablet

eBay has its own, in-house built machine translation system – eMT. Editorial credit: / Shutterstock.com

The listings were automatically translated into the language relevant to each target site. However, the whole programme backfired and resulted in a number of complaints from customers. It entailed some operational issues such as incorrect information on shipping costs populated across sites, confusion in inventory management and a substantial lack of clarity in the way sellers signed up to the programme.

Unhappy customers did not hold back when posting their feedback on social media. eBay eventually issued apologies via email.

In other news, Google Assistant went bilingual in the summer, offering extended possibilities to all multilingual households. It can now operate in two languages interchangeably, which can be particularly useful for people who live in areas with more than one official language like Quebec, where French and English are used with equal frequency.

TranslateMedia

For TranslateMedia, much like for the rest of the language industry, 2018 was about finding the most optimal ways of working with machine translation.

We strived to improve our framework and user interface to help linguists embrace this technology, very much in line with one of the central themes of the European Association of Machine Translation conference that took place in Alicante in May. The general consensus revolved around giving linguists more control and keeping them in the centre when developing machine translation technologies.

July brought the very first issue of our machine translation newsletter and in September we started providing linguists with machine translation suggestions in the Online Editor in certain languages to offer additional support and resources that could speed up their work.

No doubt there are more exciting developments just around the corner – both at TranslateMedia and the industry as a whole. It’s not going to be ‘out with the old, in with the new’, as 2019 is bound to build on the acceleration that 2018 has brought and as a result, offer new possibilities to all language professionals.

 

Related posts

Subscribe to our newsletter