In this week’s SlatorPod, we are joined by Maria Schnell to discuss her recent appointment as Chief Language Officer at RWS — pointing to the Super Agency’s investment in linguistic expertise.

Maria begins with her journey as a trained translator and how she worked her way through commercial and operational roles at SDL, before the company was acquired by RWS. She then talks about what the new role of Chief Language Officer entails, specifically leading the Language eXperience Delivery (LXD) platform.

She breaks down the role of the LXD platform, where a large, linguistic network supports clients looking to venture into more remote markets and long-tail languages. She also shares the importance of technology in optimizing operations, from translation productivity to quality assurance.

Maria discusses RWS’ approach to machine translation, where human post-editing is necessary if clients want true engagement with their end customers. She also talks about plans to simplify the technological complexity the Super Agency inherited as a result of past acquisitions.

The pod rounds off with Maria’s initiatives for the next year as she continues to nurture localization talent at RWS Campus and analyze language trends in their language labs.

Source link

Yes, machine translation technology has become surprisingly good in recent years. But that does not mean that everything is rosy in the MT camp. While the BLEU scores are improving the problem of bias has emerged. Translations produced by machines often display biases, including but not limited to gender bias.

Gender bias can happen when you are translating an expression that is gender-neutral in the source language but needs to be gender-specific in the target language, such as English doctor to German Arzt (male doctor) or Ärztin (female doctor).

If the machine is translating a sentence containing such a word from English into German, a lot depends on whether there is anything in the source text that could help the machine guess the gender of the person.

A sentence such as she is our new doctor is safe: the pronoun she is enough to nudge any well-trained machine translator toward the female translation of doctor (so, Ärztin not Arzt). But what if there are no such clues, as in I am your new doctor? How is a machine supposed to know which gender I refers to? This is an unresolvable ambiguity.

Machines usually “resolve” unresolvable ambiguities by picking whichever translation is statistically more likely, based on what they have seen more often in their training data. So doctors and directors tend to be translated as male, cleaners and caregivers as female: the translations are biased.

Gender bias affects not just nouns but also adjectives (I am happy in French is either je suis heureux if male or je suis heureuse if female); and, of course, pronouns (when translating gender-neutral pronouns from languages that have them).

Not Just About Gender

There are other kinds of biases besides gender bias. One common cause of bias in MT is the English pronoun you. Translating you into other languages often implies having to decide whether it is singular or plural, formal or informal.

If there are some clues in the source text that can signal these distinctions, then good. But if there are no such clues (as in where are you?) then again we have an unresolvable ambiguity and the machine has to make an assumption based on statistical likelihood.

The result is that machine translators tend to favor translating you as singular and formal because that is what occurs most frequently in the (mostly written and formal) texts they have been trained on. That may not be what the user intended when they typed this sentence into a machine translator, but the machine is oblivious to that.

In theory, unresolvable ambiguities can occur on any word or expression in any language pair, and you will meet them in unexpected places.

One example for many: translating river into French requires knowing whether it is a small river that flows into another river (rivière) or a large river which flows into the sea (fleuve); but this information is often not present in the source text. So, a machine translator has to make an assumption and here we have it, another biased decision has been made.

What Is Bias, Anyway?

The technical definition of bias is that it is the tendency of an automated system to make the same kind of assumptions again and again. Then there is the popular definition of bias, which is basically the same but with an implication of offense, harm, and injustice.

It is easy to see how gender bias, of all biases, can cause offense and perpetuate undesirable stereotypes. But, at its core, all MT biases are basically word-sense disambiguation problems: we have an expression in the source language that, when seen from the perspective of the target language, has two (or more) senses. The translator needs to disambiguate, to decide which one the author had in mind. Sometimes it is doable (when there is enough context to go on) and sometimes it is not. When not, then we have an unresolvable ambiguity.

Unresolvable ambiguities are different from other occasions when the MT simply gets things wrong. No AI, however smart, can ever guess what you meant if there is no trace of it in the text. This means that we cannot fix MT bias just by improving existing AI. The only way is to ask the user to disambiguate manually.

Say Hello to Human-Assisted Machine Translation

When you think about it, this realization — that we cannot fix MT bias just by improving the AI and that we have to ask users to disambiguate manually — is quite a game changer. It means we cannot treat machine translation as a linear process anymore, as a black box where in goes the text in one language and out it comes in another. The process will have to become more interactive, with humans in the loop.

Do major MT players such as Google and DeepL know this? Well, it seems that it is beginning to dawn on them. Google in particular has been taking gender bias seriously and has, as far back as 2018, launched a manual disambiguation feature for gender-specific nouns and gender-neutral pronouns in some language pairs such as English–Spanish. DeepL also has a manual disambiguation feature in some language pairs, not for gender but for changing the form of address between formal and informal.

Adding manual disambiguation into machine translation poses many challenges. First, the software needs to be able to detect that an unresolvable ambiguity has occurred. Second, it has to be able to produce alternative translations depending on the user’s choices.

Google has got to where it is now by manually annotating their training data and getting their language models to “know” about gender even when it is not overtly expressed.

DeepL is notoriously secretive but it is reasonable to assume their methods are similar.

An example of a different approach is Fairslator, a plug-in that works with any machine translator and only examines its output. A rule-based algorithm in Fairslator scans the source text for unresolvable ambiguities and another algorithm re-inflects the translation according to the user’s choices. (Disclosure: The author of this article is Fairslator’s founder.)

The bottom line is that the problem of MT bias is solvable with technology, but at a price. We will have to wean our users off the idea that MT is a black box, and bring them into the loop.

In a world where we are increasingly running out of room for further improvement in human-likeness (as measured by BLEU scores and the like), the next frontier for MT is ambiguity.

Source link

On May 18, 2022, US-based localization service provider Pixelogic Media announced the acquisition of Korean dubbing studio Juice Media.

According to Doug Higgins, SVP at Pixelogic Media, the deal was signed on May 16 and will officially close on May 31, 2022. The terms of the transaction were undisclosed.

Juice Media was founded in 2008 by CEO Minsoo Kim and specializes in delivering localization services in dubbing, subtitling, and music production. Clients include US studios, streaming platforms, and local content distributors. Based in Seoul, the company has eight employees.

Higgins told Slator that the entire Juice Media team will remain with Pixelogic following the sale and form a combined team with the existing 10 employees of Pixelogic Korea. He added, “Going forward, Juice Media will operate as Pixelogic Korea, and the company anticipates meaningful growth in the near term.”

Asked about the genesis of the deal, Higgins said Pixelogic had been working with Juice Media over the past few years to provide dubbing services for Pixelogic’s customers. He added, “As we got to know the management team, it became clear that we share a common vision focused on innovation, technology, quality, and providing solutions to customers.”

Addressing Shortfalls in Korean Dubbing Capacity

The acquisition will enhance Pixelogic’s operations in Korea by adding dubbing and audio capabilities to existing subtitling and media services. The investment will also strengthen the company’s end-to-end service offering across the localization and distribution supply chain.

“Pixelogic will continue to work with our customers to identify markets where we can help solve a misalignment between available supply vs. demand, or where we can help to address needs by leveraging our systems, tools, or end-to-end capabilities,” Higgins explained.

Embracing Remote Dubbing

Remote dubbing plays an important part of Pixelogic’s technology roadmap, according to Higgins. The company embraced remote dubbing technology at the start of the Covid-19 pandemic, using both third-party and proprietary solutions to meet clients’ needs. “Remote dubbing continues to be an extremely important requirement and we will continue to evolve our solutions for today and beyond,” he said.

On the recent increase in demand for media localization in Asia, Higgins said that Pixelogic sees this as both an “exciting and challenging” time for the company and its clients. The SVP said managing capacity requirements and timeline expectations have been their primary challenges.

Higgins also noted that the company has been very encouraged by the increase in original content being produced in Asia, saying, “It will contribute to the growth of localized non-English original content, providing a much richer offering of culturally diverse content for consumers.”

Source link

Oskar Svensson, CEO of family-run dubbing and subtitling company Eurotroll, will stay on to help French multimedia localization provider Dubbing Brothers expand into the Nordics.

The post Dubbing Brothers Expands Into Nordics, Acquires Eurotroll appeared first on Slator.

Source link

Australia data annotation provider, Appen, gets buyout offer from Telus, buyer of Lionbridge AI business. In surprising about-face, Canadian call center giant walks away.

The post Lionbridge AI Buyer, Telus, Abruptly Abandons USD 700m Appen Bid appeared first on Slator.

Source link

Prague, Czech Republic, May 26, 2022 – Memsource, the AI-powered translation management system (TMS), today announced a new translation integration for Kontent by Kentico. The integration aims to bring the power of translation management, automation, and machine translation to Kontent by Kentico, a leading headless content management system (CMS).

Memsource seamlessly integrates with Kontent to initiate, automate, and track the whole translation process. For enterprises wishing to expand to global markets and give customers a localized experience, the new integration can be set up in just a few clicks and does not require any developer support.

The integration recognizes content that needs translation, including any linked articles, and automatically launches a translation project in Memsource. When translations are complete, they are instantly sent back to Kontent, creating a truly seamless multilingual content publishing experience. Users can structure and publish multilingual experiences without leaving the Kontent platform, which is crucial for teams looking to expand their tech stack without adding bulky tools or having to learn new processes.

“We are very excited to be partnering with a leading headless CMS platform. Our partnership will allow users to leverage Kontent’s headless CMS and Memsource’s agile TMS to expand to global markets in a cost-effective way. We look forward to helping users create a seamless translation process, enabling them to publish engaging, localized experiences across all channels and customer touchpoints” said Sarma Skele, Product Manager at Memsource.

“We are delighted to partner with Memsource and to have the opportunity to offer this integration to our users. The integration will help them expand their global offering and tap into new markets with multilingual digital experiences” said Vojtech Boril, VP Growth & Marketing at Kontent by Kentico.

The integration is available for Memsource Enterprise edition and from Kontent Scale subscriptions. To start creating and publishing Multilingual content, get in touch with our team to request a demo or start a free trial.

Do you want to offer localized customer experiences to a global audience? Join our live webinar to learn more about how Kontent and Memsource can boost global growth.

About Memsource

Memsource helps global companies translate efficiently. Ranked as the the leading translation management system on G2, Memsource supports 500+ languages, 50+ file types, and 30+ machine translation engines. Memsource enables its customers to increase translation quality while reducing costs, using its patented, state-of-the-art AI technology, and serves thousands of global customers, including leading brands such as Uber, Supercell, Vistaprint, and Zendesk.

Further information is available at Follow Memsource on Twitter @Memsource.

About Kontent by Kentico

Enable marketers. Free up developers. Unify teams. Kontent by Kentico is the headless CMS where modern digital experiences are made. In the content hub, marketers plan and publish resonating content, while developers make the experience look and feel great on any channel. Seamless, governed workflows and control meet independence and agility. This is unified content management.

Media contact:
Allison Spangler, Product Marketing Manager

Source link

Deep down, every business person in the world knows it: localizing a company’s content and products can boost revenue — enormously. But how can globalization experts persuade the C-suite — and other stakeholders — to greenlight a localization program?
In other words, how do you pitch a localization program? Answer: you create a business case.

The post Want to Win That Localization Project? Build a Better Business Case! appeared first on Nimdzi.

Source link

A coalition of Belgian linguist groups has published a guide to help public bodies that purchase interpretation services improve their tender specifications.

The foundation for the 23-page vade mecum was laid back in October 2021, when Freelance Interpreters United (FIU) and the Belgian Chamber of Translators and Interpreters (CBTI-BKVT) presented a “Charter of Good Practice and Collegiate for Interpreters” at a seminar in Brussels.

“This charter is an important first step towards optimal working conditions, widely supported by our interpreting colleagues and aimed at ensuring a quality service,” CBTI-BKVT wrote in a Facebook post. “Together we can shape our profession.”

The Belgian Quality Translation Association (BQTA) then joined the two groups in putting together the new guide for authors of tender specifications. The manual references the charter as a resource to help set realistic demands for interpretation services.

The vade mecum also defines commonly used industry terms and explains the differences between freelancers and subcontractors. It devotes a number of pages to distinguishing simultaneous and consecutive interpreting, from tech to the number of interpreters needed.

When it comes to tender specifications, the manual recommends using a standardized tender form with clear and complete specifications (e.g., setup, equipment required, as well as including precise information on the number of meetings, dates, venues, and language combinations.

In addition, the guide acknowledges that, in some situations, open-ended questions are appropriate. This allows experienced candidates to describe their own proposals in response to the buyer’s needs.

Tough Decisions

Selection criteria feature heavily in the guide, with the authors touching on qualifications to consider, codes of conduct, and quality management. Interestingly enough, the factors included in economic criteria vary from the standard (solvency) to the more advocacy-oriented (payment terms and remuneration for subcontractors, corporate social responsibility).  

“The role of governments in promoting fair and sustainable business practices should also be highlighted here,” the guide stated. “Providing opportunities for local, self-employed interpreters or SMEs will also strengthen and sustain the much-needed national economic fabric.” 

The vade mecum offers several “alternatives to ‘winner-take-all’ contracts” that alternate among several contractors; using a cascade system (i.e., ranking contractors by score), and dividing a contract into lots to be assigned to one or more contractors.

This last option, also known as a framework agreement, has become increasingly popular among government agencies in a handful of European countries. Results have been mixed.

Interpreters in Finland, for instance, staged a three-minute work stoppage in November 2021 to protest issues that could be tied to framework agreements: interpreters’ limited negotiating power and a perceived hyperfocus on price in the government’s tender process.

Price should be given a maximum weight of 50% during the evaluation process, according to the vade mecum. The guide recommends paying “particular attention to quality, methodology, approach to technical support, professional experience, useful references, and local anchorage.” 

Moreover, buyers should take into account cost-of-living increases for multi-year contracts and charges for recording interpretation — a new challenge interpreters have faced since remote simultaneous interpreting (RSI) became the de facto interpreting mode during pandemic lockdowns.

The guide takes the stance that recording or broadcasting interpretation requires prior consent from the interpreters providing the service.

“The interpreting service is exclusively intended for immediate and direct use by the audience during the interpreting assignment and should be considered a communication aid,” the authors wrote. “In the case where the recording is broadcast or distributed, the market practice is to apply a supplement of at least 30% of the fee.”

Source link

Slator is the leading source of news and research for the global translation, localization, and language technology industry. Our Advisory practice is a trusted partner to clients looking for independent analysis. Headquartered in Zurich, Slator has a presence in Asia, Europe, and the US.

Source link

With 70% of respondents reporting that less than half of their customer touchpoints are multilingual, the Global Experience is fragmented and in need of more enterprise support

San Francisco, CA – May 24, 2022 – Lilt, the leading global experience platform enabling organizations to build and deliver multilingual experiences, today announced the results of their 2022 State of Global Experience report. With over 1,800 responses from various departments, companies, and geographies, the survey and report aim to capture the strategies, pain points, and best practices of global customer experience and engagement.

“Global customers embark on a journey made up of countless touchpoints, each offering personalization and instant access at any time. These digital stepping stones, providing the opportunity to engage consistently and hyper-locally, have fundamentally changed what customers expect of businesses,” said Paula (Barbary) Shannon, Chief Evangelist at Lilt. “The 2022 report highlights how language of choice is increasingly becoming a baseline customer expectation, and customer-focused businesses are paying attention.”

Key insights from the report highlight:

  • A fragmented customer experience: Just 12% of respondents reported their digital customer touchpoints are fully multilingual in the languages their company supports, with more than 70% reporting that less than half of their touchpoints are multilingual.
  • More room for consistency: 42% of respondents reported that their company doesn’t localize in all countries where they have a presence.
  • A strategy and execution gap: While nearly 90% of respondents consider global experience (GX) a priority within their companies’ target markets, 40% don’t believe their company effectively manages it.
  • Lack of GX alignment and cohesive messaging: Despite more than two-thirds of respondents reporting that their company understands the importance of personalization throughout the customer journey, 62% note that their company doesn’t have alignment on global experience (GX) strategies, and 65% don’t have cohesive messaging across locales and markets. 

The Transition To Digital Continues

As the world goes digital, consumers are exposed to more companies and products than ever before. In this new world, language becomes an increasingly important part of how they engage and communicate. Companies have begun to understand the value of language in customer experience. Lilt’s survey found that 95% of respondents believe language of choice plays an important role throughout the customer journey.

Lilt has found that leading companies are taking notice of this trend, and investing more resources into elevating their global customer journeys, with 63% of respondents reporting growing their volume of multilingual content over the past three years. As the world continues to move digital-first, respondents also shared anticipation for a continued increase in demand for multilingual content..

The Future of Global Experience

For companies to elevate their global customer journeys and provide more localized experiences, they will not only have to grow their content volumes, but also ensure that GX  is woven into cross-functional company strategy. Three key areas were identified by leading companies as priorities in the near future, including:

  • Strategic Resource Allocation: Many will focus on accessing the right resources at the right time. While funding is an ever-present need, many also highlighted the importance of dedicating more time and strategic thinking to global experience management.
  • Embodying a Customer-Centric Approach Globally: Many respondents also highlighted a proactive effort to operate more cross-functionally, see things from the customer perspective, and deliver one cohesive experience for every customer around the world.
  • Automation and Technology: As volumes of content continue to grow, numerous companies noted a focus on identifying and implementing technology applications that can automate processes, enabling teams to accomplish more with the same resources and headcount.

Additional information and insights are available in the report: 2022 State of Global Experience Report. For more information about the report or Lilt and its translation solutions, please reach out to

About Lilt

Headquartered in San Francisco, Lilt is the leading global experience platform, enabling organizations to build and deliver multilingual experiences at scale across every step of the global customer journey through its translation technology and services.  The Lilt Platform uses AI and automation to make the localization process faster, better, and simpler, bringing human-powered, technology-assisted translations to global enterprises. Lilt gives industry-leading organizations like Intel, ASICS, Emerson, UIPath, and Canva everything they need to scale their global experience programs and go-to-market faster. Lilt has additional global offices across North America, Europe, and Asia. Visit us online at or contact us at

Kat Eller Murray

ROAM Communications for Lilt


Source link