Discover the Latest Trends in the Translation and Localization Industry

Discover the Latest Trends in the Translation and Localization Industry

Nov 29, 2022 – Berlin, Germany – It may appear that in 2022, with the level of quality achievable by Machine Translation (MT), localization firms providing human-based translation and editing are no longer required. However, customer demand for localized content is increasing across the world, and the sector is on the rise rather than the decline. Let’s see what stands behind the increasing demand for translation at the end of 2022.

Neural Networks – the driving force of the translation industry

The introduction of neural machine translation (NMT) in 2015 resulted in a considerable improvement in machine translation quality. NMT, which is based on artificial neural networks, imitates the activity of the human brain in integrating and encoding information and proposes a technique capable of reducing gender and case bias in MT.

A neural MT system, unlike a statistical MT system, does not contain minor sub-components that have to be calibrated independently. Instead, it creates a large network of interconnected components, which the system continually fine-tunes as it is utilized. Neural networks and deep learning are now the driving forces behind cutting-edge MT algorithms.

Marketplaces and Airbnb listings: in what industries does raw machine translation work?

According to a report released in 2020 by Google Deep Learning researchers, NMT is on the verge of beating human-level translation, which suggests that the final quality machine translation can give is almost there. However, the machine translation quality provided by tech giants such as Google Translate, Amazon, Microsoft, and other major firms is not expected to improve significantly over the coming years.

Roman Kotzsch, CEO at Milengo, mentions that while customers don’t mind utilizing MT solutions for personal purposes, when it comes to corporate translation, quality standards skyrocket.

“When translating a personal email or a chat, consumers are satisfied with the quality of Google Translate or DeepL. However, when it comes to an app or a commercial website, the expectation for consistency, terminology, and error-free translation becomes considerably greater as these factors contribute towards brand reputation and consumer trust. As a result, if a company’s translations are of poor quality, it will immediately undermine the brand.”

However, there are certain business scenarios where “raw” machine translation is sufficient. E-commerce is a fantastic example: online stores upload new items into their catalogues every week, and often these products are only available for a few weeks. The product description has a short lifecycle in this scenario, and machine translation works well enough for this. Companies must, however, modify MT engines for specific sectors and content types.

Reviews and descriptions on marketplaces such as Airbnb are other examples of content with short lifecycles. In fact, a study of Airbnb’s top ten languages by a machine translation evaluation company showed that their Translation Engine enhanced the quality of more than 99% of the platform’s listings.

Video content and data annotation: who is the next king?

Localization of audio and video content, particularly AI-based voiceover, is currently one of the industry’s most important drivers. Over 80% of total internet traffic is driven by global online video platforms. As preferred content shifts from text to audio and video, the younger generation is less likely to read. As a result, many companies today offer a choice of written books, audio guides, and video courses on the same topic to their customers.

Data annotation is the process of labeling data from text, photos, or videos for AI systems. It is another significant development driver of the translation sector. Major participants in the localization market, such as Transperfect, RWS Holdings, Appen, and others, provide data annotation and labeling as a separate service. There are two reasons why this has become a trend in the localization market: firstly, because data annotation is culturally specific, and thus cannot be totally automated. To put it another way, the same words, pictures, and things might be classified differently in various cultures. The classic examples with “football” and “soccer” or “chips” and “French fries” are just a top of this iceberg. Secondly, language service providers are particularly effective at sourcing freelance personnel on a global scale, making the service easier and less expensive for the consumer.

Human expertise is still crucial

The translation industry’s latest breakthroughs are exciting and encouraging. Milengo is especially interested in watching how machine translation’s potential improves beyond 2022. However, until it becomes a completely dependable tool, it is critical to invest in quality translations or at minimum, post-editing from skilled translators to ensure that your corporate content properly satisfies the demands of ever-expanding worldwide markets.

Looking for a reliable translation service provider? Look no further! Milengo makes localization easy and affordable with 30 years of industry experience. Our network of top industry specialists, including over 1,200 translators/interpreters/voice artists/transcribers worldwide, will guide you to your objectives with their knowledge and skills.

Jane Doe

read all posts

Jane is The Recursive’s Western Balkans Editor, covering tech, innovation, and business for more than a decade. He’s currently exploring blockchain, Industry 4.0, AI, and is always open to covering diverse and exciting topics in the Western Balkans countries. His work has been featured in global media outlets such as Foreign Policy, WSJ, ZDNet, and Balkan Insight.