Live Chat

Higher quality machine translations, new languages and productivity enhancements

Introducing the industries' next generation of machine translation technology - XMT. XMT is a new modular, flexible and adaptive translation technology powered by state-of-the-art machine learning algorithms to improve translation quality and user experience. It enables language-specific enhancements and allows user feedback to be incorporated seamlessly and in real-time leading to a continuous translation quality improvement. The combination of advanced algorithms, additional user control and flexibility leads to:

  • Improved MT quality: higher quality machine translations, particularly for challenging language pairs such as English to Asian languages
  • Support for new languages and content formats: Translate texts, tweets and other informal communication formats accurately into the target language
  • Future productivity enhancements: XMT provides a platform for new MT capabilities and translation productivity functions, including language learning

Translate informal languages

Unique algorithms have been developed to support texts, tweets and other informal communications. For example, XMT understands the abbreviations and short hand used to fit in the 140 character limit of Twitter. As new communication formats emerge, new algorithms can be applied without disrupting those already working.

Higher quality translations

Current machine translation technologies apply a single translation algorithm universally to all language pairs. XMT allows different translation algorithms to be used for different language pairs so the best can be applied to each language. This is particularly important for accurate translations of Asian languages, which has historically been difficult for machine translation technology. The result is high quality translations, regardless of language.

New MT productivity tools

XMT is a complete rewrite of MT technology by a group who understands the limitations of previous generations of machine translation. The XMT architecture forms the basis for future machine translation developments from SDL. One of the new capabilities will be Language Learning, the ability for MT engines to learn user preferences. This capability will be released later this year. 

Check out how this works. 

Translation engine training

Similar to previous generations of MT, XMT based-engines can be pre-trained with industry or company specific vocabulary or style preferences. SDL XMT engines are available pre-trained with vocabulary from the automotive, electronics, IT, life sciences and travel industries, with new industries being added over time.