Translution logo


FAQs - Translation Quality

Where can I get information on improving translation quality?

Please click here to review our white paper: Predictions on the Future of the Translation Industry. As well as discussing the developments that are happening in the translation industry, this also explains how machine translation can be improved

You may also want to review our Translation Tips, which provide a "primer"

We also run comprehensive training courses which include modules on "Writing for Machine Translation". Please click here for further details.

Back to top

What tools to improve quality are included with Translution's products?

Translution Pro and Translution Business come with a range of tools built in which improve the translation quality including:

  • Dictionaries for specialist terminology
  • Marking text as Do Not Translate
  • Dictionaries of Do Not Translate (DNT) terms such as names, brands, places etc.
  • Tips on how to write text to improve the translation quality.
  • Use of Translution's own dictionary of Not Found Words and common expressions
  • 20 Industry specific/ specialized dictionaries

User Dictionaries

Everyone uses words and phrases that, to them, have particular meanings. For example: parts lists, where a specific translation is required for an individual part description. All Translution users can set up their own User Dictionaries, which control how the translation is handled.

Shared Dictionaries

Users of our business products (Translution Business, Translution Web, Translution API and Translution Workbench) can share dictionaries across their entire organisation. They can also develop their own Domain dictionaries (for specialist subject areas or departments).

"Do Not Translate" lists

Translution provides a facility for the user to mark up text within a specific document that they do not want translated. It also provides Do Not Translate lists i.e. words which the user or the organisation would never want translated. Recent research undertaken by Professor Tony Hartley of the University of Leeds, shows that implementing Do Not Translate lists improves local translation accuracy by, on average, 23%. (Ref: Improving Machine Translation Quality with Automatic Named Entity Recognition, EAMT Workshop at EACL 2003 Budapest 13 April 2003).

Managing Already Translated Text

It is essential that already translated text is not re-translated as this reduces overall translation quality. This is important, particularly when replying to or forwarding emails across language barriers. Translution ensures this does not happen by always including the original text with the translated text, and by identifying already translated text in email replies and forwarded messages. Translution also ensures already translated web pages and documents are not re-translated.

Handling "Not Found Words"

Machine Translation (MT) engines are usually supplied with dictionaries of the most common words used in a language. However no MT engine could possibly include all the words in a language, particularly "rich" languages such as English, French and German. Until these words are translated and added to the dictionaries of MT engines held centrally, they will not be recognized by the translation engines and will simply pass through un-translated. As part of its ongoing commitment to improve translation quality, Translution collects and analyses all Not Found Words generated by all users and, after researching them, continuously adds selected new words and phrases to its dictionaries. We also add common idiomatic phrases and their translations. In this way overall quality of translation is improved for all users of Translution products.

Specialized Dictionaries

We provide 20 specialized dictionaries covering the following areas which can be used automatically by selecting a dictionary profile. Easy to set up and used in conjunction with your own dictionaries these can have a dramatic impact on the accuracy of translation.

Dictionaries are organized into 6 groups: Business, Colloquial, High Tech, Industrial, Military and Science.

The individual dictionaries supported include: Automotive, Aviation/Space, Chemistry, Colloquial, Computers/Data Processing, Earth Sciences, Economics/Business, Electronics, Food Sciences, Legal, Life Sciences, Mathematics, Mechanical Engineering, Medicine, Metallurgy, Military Science, Naval/Maritime, Photography/Optics, Physics/Atomic Energy, Political Science.

Dictionary Profiles

Organizations using our business products (Translution Business, Translution Web, Translution Workbench and Translution API) can use up to 12 dictionaries at any time organized in order of priority.

Users of Translution Pro - our product for individuals can set up only 1 dictionary. However this can be used in conjunction with our specialized dictionaries.

We therefore enable users to set up a dictionary profile - a bit like setting up a style sheet - which when selected applies your rules of priority and which dictionaries are used automatically

Back to top

Do you have any tips on how to write for translation?

Please click here to review our Translation Tips.

Back to top

What is Translation Memory?

Translation Memory (TM) is source and target human translations stored and aligned into segments (usually sentences) in a database. Using TM for both machine and human translation ensures consistent and perfect translations of "standard phrases".

Source documents are scanned at Translution Central and compared with the translation memory database before they are presented to the translation engines. If a 100% match Translation memory is found this is used in preference to the machine translation.

Storing, managing and using Translation Memory also can dramatically reduce the cost of translation by Translution's human translation teams as we offer discounts for "fuzzy matches"

All Translution's products (except Translution Pro) use Translation Memory.

Back to top

Does Machine Translation remove the need for expensive human translation?

For day to day communication and comprehension, Machine Translation does indeed remove the need for expensive human translation. Where translation accuracy must be assured i.e. for legal documents and marketing material and websites, human translation is still advised.

For further information on where you can use machine translation and where you should use human translation click here

Back to top

Can the software handle idiomatic phrases?

Users can add idiomatic phrases that they use regularly to their dictionaries as fixed expressions, although we advise that you avoid using idiomatic expressions when you write as, without the use of a dictionary entry, the translations tend to be poor. We are also adding common idiomatic translations into Translution's dictionaries for the benefit of all our users.

Back to top

Which Machine Translation Engines are Translution using to provide translations?

Initially we are licensing machine translation engines from one of the leading companies in this sector (Systran). We will soon be integrating other machine translation engines to grow the number of languages we support.

We have also carefully evaluated six companies' machine translation engines and others are planned. Click here for more details.

Translution uses an "open architecture" and is "agnostic" to the machine translation engine used.

The benefits of having an "open architecture" are as follows:

  • Our users will eventually be able to translate across a much wider range of languages than any other machine translation provider
  • Improvements in machine translation quality can be made available as they occur
  • We can offer the best machine translation for individual language pairs or even for individual language directions.

Back to top