I attended a recent TAUS Asia Conference and Exhibits event, which was held in Singapore from 16-18 October 2019. TAUS is a language data network who describe themselves as “an independent and neutral industry organization” seeking to develop communities by sharing knowledge, metrics and data during events and in online user groups. They’ve organised 25 events in Asia since 2005, but this was the first time they’d staged one in Singapore.
It was the largest event ever organised by TAUS, drawing in over 210 individuals in the translation and interpreting industry from 22 countries around the globe. Attendees at the event were encouraged to learn how to redesign new translation processes, deepen their knowledge of emerging technologies, and meet new customers and vendors in the language field.
As with all TAUS conferences, the information presented was at a high level, engaging and presented by top industry experts. I learned about the latest developments in relevant technologies particularly current best practices in the continuously-evolving and rapidly-advancing translation and interpreting industry.
Topics at the event were fascinating and varied. Some of the most prominent discussions around global content included (among others): ensuring translation quality and how to deal with emerging markets; how companies such as Google, Microsoft and Amazon tackle language industry issues and solve customer challenges; the state of neural machine translation (NMT); the knowledge, operation, and data gaps in the industry; solving localization challenges with the modern translation pipeline; and Artificial Intelligence to automate subtitle creation and translation (this was one of the most interesting presentations in the whole conference).
No language industry conference would be complete without numerous presentations on artificial intelligence and Neural Machine Translation (NMT), and the TAUS conference in Singapore was no exception! It is common place for language companies in the USA and Europe to use NMT to solve client translation needs. Attending conferences like those organised by TAUS reinforced my understanding of the global businesses' wide-spread use of NMT today.
Presenters discussed the benefits of having robust NMT systems that produce context-appropriate translations and limit human involvement. Machine translation is improving every day and new developments are constantly rolled out, so it’s imperative that translators and language companies stay on top of the latest trends and innovations. Attending conferences like TAUS that feature leading minds in the field are fantastic opportunities for companies and individuals to learn even more about the latest developments in the industry.
By attending the TAUS conference, I was able to further increase my knowledge on the importance of data, domain-sourcing, and methods of evaluation for MT quality. MT systems developers test daily to see if new data and algorithms lead to improvements in output quality. Due to the sheer number of tests, human evaluation simply isn’t feasible, so developers use automated metrics instead. The most commonly used metric is BLEU (Bilingual Evaluation Understudy).
The fact that machines are completing a significant number of translation text today should not be a concern to most translators. The industry is growing and there is a significant amount of work, increasing every day and available to translators who use technology.
Machines will not replace translators. The translators who adapt with the new technologies will replace the translators who do not use the new technologies.
One of the highlights of the TAUS conference every year is a contest in which participants present their game-changing innovations. Ten companies took part in the 2019 competition, including NICT, ModernMT, Transn, Auto Desk and Xillio, with innovations in language processing and linguistic quality management among the ideas featured.
This year’s winner was Adam Hodgson, founder and CEO of Interpreter.io. The game-changing technology presented was the use of Artificial Intelligence to assist interpreters (termed AI AI). Interperter.io technology provides automatic interpretation in real-time and is used to help interpreters with their interpretation as they refer to the real-time translation. For more information about the event refer here.
We highly recommend attending a TAUS Conference, whether one pops up in your city or whether you travel to attend. The information presented is well-worth the time and investment. The next TAUS conference is the TAUS Global Content Conference & Exhibits in San Jose scheduled for 10-11 March 2020. The main theme at this event is Powering Language Data. For more information contact here. I look forward to seeing you there.