Team-up between Middlebury and Irish machine translation startup KantanMT might herald next step in immersive language learning
Share on Facebook
Share on Twitter
Share on Google+
Share on Reddit
Share on Email

Middlebury (image via Middlebury on Twitter)

The partnership will mean students studying advanced translation will get access to top-notch tools

Irish machine translation company KantanMT has reached a partnership agreement with the immersive language school Middlebury Institute of International Studies at Monterey, in what could herald a demonstrable upgrade in language-learning programs. The partnership will apply to Middlebury’s Advanced CAT (Computer Aided Translation) course, taught by Assistant Professors Max Troyer (TLM Program Coordinator) and Adam Wooten.

Middlebury is famous for its intensive, two-to-three-month long intensive language programs. KantanMT is in the “localization” business, translating websites, apps, and platforms into other languages or local dialects. Utilizing machine translation (MT) should give Middlebury’s portfolio of programs a strong pick-me-up.

They are known for their demand that students sign a “Language Pledge” only to use the language of study as their primary method of communication throughout the duration of a school’s program (and of course in real emergencies you can default to your native language)

“KantanMT.com provides the leading cloud-based Machine Translation platform for the localization industry. It offers everything needed to integrate Machine Translation technology into corporate communication channels and localization workflows,” Troyer wrote in a press release. “No localization training program is complete without the latest in machine translation technology, and the Translation and Localization Management program at MIIS is no exception.”

Kantan is one of several companies that offers “translation memory” banks to clients that speed up the process of interpreting documents, blueprints, and software based on the software program’s ‘memory’ of previous translation projects. Put another way, translation memory (TM) builds up a vocabulary of keywords and phrases over time to assist companies with complex, specialized translations (say English documents on US constitutional law into Arabic or engineering blueprints from Japanese into French).

KantanMT (screenshot)

KantanMT sells customizable and pre-stocked translation memories to clients in a handful of industries, including medical, automotive, legal, financial, and technical. They were a finalist in The Irish Times’ Innovation Awards 2016.

“The partnership between KantanMT and MIIS TLM ensures our curriculum will include the latest developments in the translation technology industry,” Troyer continued.

The school’s MT program will cover English, French, Brazilian Portuguese, Mandarin Chinese, Taiwanese Mandarin, Korean, Japanese, Russian, Spanish, and German. Their immersion programs also cover Arabic and Hebrew. About 100 students will use KantanMT’s signature software as well as KantanLQR (language quality review) to go over their own translation skills.

“With rapidly advancing translation technologies and increasingly more use cases for machine translation, our graduates must understand not only how to train and customize a machine translation engine, but also why they are doing it the way they do,” according to Wooten. “In our translation technology courses at MIIS, students learn why they use the technology a certain way so they can make better decisions even as new technologies and uses continue to evolve.”

“For this reason, we appreciate that KantanMT not only provides students with access to its technology but also publishes resources and insights that explain the reasons their technology was engineered to work the way it does.”

Having translation tools like KantanMT or even the open Google Translate available can, in the simplest ways, be used to check the performance of a language student. Several studies, like this one using MT engine Tradukka and another using Hebrew-English Targumatik, show this strategy is instrumental in learning new languages. It is easy to see how a tool like this could be valuable to people learning languages as well, not just translating them.

Led by Founder and CEO Tony O’Dowd and CCO Bryan Coyle. Delta Partners’ John Kenny is on the company board.

Share on:Share
Share on Facebook
Share on Twitter
Share on Google+
Share on Reddit
Share on Email

More Goodies From Big Data


How Cognitive Search Eliminates Common Struggles Website Users Face

How did Big Data transform the manufacturing industry?

10 ways to save money with AWS Redshift