Recent MT Trends, Research, and Features Taking Companies Global

Localization is a rapidly growing and evolving industry, enabling companies to go global through new methods and technologies. On top of this, new research in the field is helping bring to light the benefits of AI, machine learning, and other areas that are contributing to the increased accuracy and efficiency of one area in particular — machine translation (MT).

Here are the top recent trends and research in the localization and MT worlds, and must-reads for companies as they continue to go global.

The intelligent translation era: the next step in going global

“When trying to reach a global audience, the words themselves are cheaper than the context.” UK Tech details the importance of localization, how global grands are using it to reach their growing audiences, and how this process has become more efficient and accurate for marketers thanks to today’s technology enabling the merging of machine learning and human optimization.

Facebook funds AI and MT research to solve challenges in natural language processing

In an effort to improve machine translation of non-English languages, Facebook has announced the launch of a new initiative, providing funds for researchers to investigate ways to improve natural language processing and support neural machine translation research, with a focus on how to use the emerging technologies to translate different languages into English.

Emulator uses AI to offer the translations your games never had

While game localization is gaining speed, many classic video games are still only accessible in a single language. Engadget explores how a new software, Version 1.7.8 of the RetroArch emulator, will help gamers worldwide enjoy their favorite games through a new “AI Service feature that uses machine learning to translate game text into the language of your choice” by tapping into services like Google.

Study finds that Machine Translation could be more easily used for subtitling

As far as content localization goes, subtitling remains one of the last types that machine translation has yet to crack, since it is “highly nuanced [and] made up of dialog that contains colloquialisms, cultural references, and humor.” This recent article by Slator highlights new research that reveals a “strong case for machine subtitling in translation” and how it can be done with new technological findings.

How deep learning is impacting the translation industry

In the 1950s, leading researchers thought it would just be a few years before they were able to crack the machine translation puzzle. While the prediction was off, recent MT research reveals that deep learning could be the key that cracks the machine translation code. Aithority analyzes how human translation and the concept of a translation agency as we understand it today could transform as a result.