The future of machine translation and how to use it properly

CEO and founder of Lingohub. Envisioning a multilingual digital world. Email me if you have questions about how Lingohub can help you take your products global.

Helmut Industry Updates 0 Comments

SHARE

The Future of Machine Translation and how to use it properly

Lots of common myths about Machine Translation exist, including that it’s about to replace human translators. This blog post aims to break the myths and show how both businesses and translators can benefit from Machine Translation systems to accelerate translation.

What is Machine Translation?

Machine Translation (MT), also referred to as automatic or instant translation, is the automated translation of text by a computer. It provides text translations based on computer algorithms without human involvement. Generally there are two types of Machine Translation to distinguish:

Rule-based Machine Translation (RBMT) Statistical Machine Translation (SMT)
Data relies on linguistic rules and bilingual dictionaries for every language pair instead of extensive databases, it uses statistical translation models from analyzing training data
Translation uses rule sets to transfer the grammatical structure of source content into the target language is selected from training data with algorithms to select the most commonly appearing words
Costs for creating MT initial and ongoing investment to increase translation quality steadily extensive hardware configuration is necessary to run Machine Translation

 

How to use Machine Translation the right way

If you use Machine Translation the wrong way it might harm your business because grammatically incorrect and faulty translations are likely to lead to bad user experience. These 4 use cases help your business and translators to profit from Machine Translation:

  •  Use Machine Translation if no Translation Memory suggestions are available
    A Translation Memory suggests translations (exact or fuzzy matches) based on previously translated texts. If a Translation Memory lacks of suggestions for a text, use Machine Translation as an initial phrase in order to work out high-quality translations.
  • Machine Translation for workflow optimization
    If used correctly Machine Translation will accelerate your translation workflow. Human translators then edit Machine Translation output (post-editing) to create high-quality human translations.
    LingoHub enables you to automate this workflow because lots of translation work is processed by the system without assistance. The powerful autofill feature automatically fills up empty text segments with suggestions from MT. Human translators just need to post-edit translations, thus, it ultimately speeds up the translation workflow.
  • Machine Translation for user generated content
    Frequent changes and updates are characteristics for user generated content. Translating all user generated content by professional translators will be expensive. Machine Translation is a less expensive and quite reliable alternative according to data from GALA Global:

    • 5-20% of Machine Translation suggestions can be used as final translations
    • Roughly 40% of suggestions can be published after post-editing
    • MT provides data to autocomplete for up to 80% of texts
  • Use Machine Translation for English, French, Spanish and Portuguese texts and translations
    Machine Translation engines best translate text from and into English, French, Spanish and Portuguese. On the contrary, Russian, Polish and Korean have lower leverage rates with 5% complete matches.

The future of machine translation and how to use it properly

The Future of Machine Translation

Machine Translation has already become part of our everyday life. Millions of people use Google Translate every day. Integrated in the Google’s Chrome browser Google Translate translates a website whenever you come to a page written in a language you don’t understand this one time.

Ten years after announcing the launch of Google Translate, the U.S. based company has now announced the Google Neural Machine Translation system (GNMT). The significant difference to Phrase-Based Machine Translation is that it does no longer split the input sentence (source text) into words and phrases before translation. Neural Machine Translation regards the entire source text (sentence) as a unit for translation.

Summing it up

As Google states Machine Translation does not yet keep up with human translators, however GNMT is an important step in the right direction to enhance Machine Translation’s capabilities. Using Machine Translation the right way will accelerate your translation speed, minimize per-word costs and relieve your translators. Try it for yourself and start your free LingoHub trial today!