Blog

MT Summit 2019: Gathering Machine Translation Stakeholders Under One Roof

machine translation summit 2019 localization events

The 17th Machine Translation Summit took place in Dublin last week from 19 - 23 August 2019. From research to river dancing, the event was jam-packed with interesting talks and fun events including valuable insights from MT (machine translation) experts, and great debates fuelled by diverse attendees. It provided an opportunity to deep-dive into the world of MT and learn about the latest research and development projects that are going on.

To kick off the MT Summit there were two days of workshops and tutorials, followed by a three-day conference program consisting of four tracks: Research, Users, Translators, and Projects. The Research track covered a variety of topics including domain adaptation of MT, speech-to-text translation, and post-editing. The award for the Summit’s best overall presentation went to a presenter from this track, Antonio Toral for his research comparing human translation and post-editing. The User track presentations consisted of real-world user scenarios reflecting just how prevalent MT is in the commercial space. In this track, topics focused on domain-specific MT and neural post-editing and quality estimation. The more interactive poster sessions covered all tracks and gave attendees the opportunity to engage with the presenters and peers.

A notable and welcomed addition to this year’s MT Summit was the Translators track. Traditionally, the Summit has focused on MT researchers, users, and vendors. The Translator track has added a new aspect to the already inclusive conference and prompted lively discussions during the event. We interviewed Celia Rico and Federico Gaspari, the Translator track co-chairs, to find out how it went, what translators can expect from these MT Summits, and why they should attend MT Summit 2021.

Why was a dedicated Translator track added this year?

This exciting development built on the success of the 21st annual EAMT (European Association for Machine Translation) Conference, held in Alicante, Spain, roughly a year ago. This event featured a dedicated Translator track which added an important new dimension to the multi-faceted contemporary debate on MT.

What were the aims of the new track in the MT Summit program?

The aim of the novel Translator track was to involve translators into the MT conversation on an equal footing with researchers, developers, vendors, and users of translation technologies; the conference organizers felt that it was high time for their often neglected voice to be heard loud and clear. With this objective in mind, specific topics of interest included in the track concerned issues that increasingly confront language and translation professionals on a daily basis. They included productivity measurements and their impact on MT adoption, the role of MT in translators’ work (pricing issues, post-editing tasks assignment and their acceptance among professionals), ethical and confidentiality considerations when using MT, and psycho-social aspects of MT adoption, such as attitudes and (pre-)conceptions.

Why should translators attend the next MT Summit?

This is a really exciting time for translation technologies, and machine translation in particular, especially due to the rise of the very promising neural approach, which shows substantial improvements over the previous statistical paradigm. As with any technology, the development of translation tools will be shaped according to the needs and requirements of those involved in the relevant development processes. If development decisions are left entirely up to computer scientists, programmers, engineers, and business leaders, most affected by machine translation. We strongly believe that involving translators in this debate, especially at a traditionally inclusive and well-attended international conference like the MT Summit, increases translators’ awareness of the actual potential and inherent flaws of these much-talked-about, but often little understood, technologies, and in turn benefits their longer-term development according to the real needs and expectations of translation professionals.

How successful was the Translator track?

The Translator track brought together a representative group of translators, coming from institutional organizations, the industry, and the freelance world. They had the opportunity to express their views and concerns on the different issues when putting MT to practical use: how translation processes are affected, whether the translator needs to adopt a new role as a post-editor, or what the pressure is on prices and productivity. These are all questions that have certainly been discussed before, but the novelty this time was having the opportunity to enter the debate with other MT stakeholders. The large room that was used for the track was full to capacity for both sessions of oral presentations over two consecutive days, which in itself is an indication of the interest in the track. Similarly, the papers accepted for the poster presentation in the Translator track attracted significant interest and interactions with the conference delegates. This initiative can only enrich the MT discussion, and, as the Translator track organizers, we feel this timely event has made a significant contribution to promoting the inclusion of translators in the debate on MT, which will hopefully remain the case in the future.

CONCLUSION

This year’s MT Summit was a great opportunity for all MT stakeholders to get together to discuss and debate the challenges and possibilities of MT, new developments, and also network with industry peers both at the conference and at the social events. The next MT Summit will take place in 2021 in Seattle. But if you can’t wait that long, you can also attend the EAMT annual conference in Lisbon next year.

If you want to learn more about the research presented at the conference, you can check out the MT Summit 2019 proceedings.

Special thanks to Celia Rico and Federico Gaspari for sharing their insights.

Review MT quality scores before any post-editing is done.

Learn more
Author: Lara-Ashleigh Pieterse
Author: Lara-Ashleigh Pieterse

Content Manager at Memsource