Machine translation
Thanks to the technological advances of the last decade, machine translation has become a real alternative to human translation, offering professional results and significant cost savings for many types of texts.
In addition to the use of general machine translation engines, we can also create custom machine translation engines for our clients upon request, using their previously translated material supplemented with other large collections (corpora) of texts in the given subject or technical field, which we update and refine on a regular basis to ensure the best possible machine translation quality.
Post-editing
Depending on the intended use of the texts, a human post-editing step should be added to the general internal quality control. In terms of post-editing workflows, we distinguish between light and full post-editing.
Light post-editing
The aim of light post-editing is to produce grammatically and semantically correct, error-free, and easy-to-understand text from the machine-translated content quickly with minimal human intervention.
Full post-editing
In the case of full post-editing, the final text meets all criteria for human translation.
When should you consider machine translation and post-editing services?
- if turnaround time is of primary importance
- if the available budget does not allow for human translation
- if the primary goal is to understand the gist of the text
- when large amounts of text need to be translated
- if you need to translate a large number of documents with similar, repetitive linguistic elements
Machine translation with a general translation engine + light post-editing | Machine translation with a general translation engine + full post-editing | Machine translation with a custom translation engine + light post-editing | Machine translation with a custom translation engine + full post-editing | |
Quality | ★★☆☆☆ | ★★★★☆ | ★★★☆☆ | ★★★★★ |
Turnaround time | ★★☆☆☆ | ★★★☆☆ | ★★☆☆☆ | ★★★☆☆ |
Price | ★☆☆☆☆ | ★★☆☆☆ | ★★☆☆☆ | ★★★☆☆ |
Workflow steps | ||||
Format check | ✓ | ✓ | ✓ | ✓ |
Basic quality control and consistency checks | ✓ | ✓ | ✓ | ✓ |
Terminology check | correction of critical terms | full terminology correction | correction of critical terms | full terminology correction |
Spell-check | correction of critical errors | correction of all errors | correction of critical errors | correction of all errors |
Bilingual content and grammar check | correction of critical errors | correction of all errors | correction of critical errors | correction of all errors |
Style check | X | ✓ | X | ✓ |
When to use | ||||
Goal | To produce a grammatically and semantically correct, easy-to-understand translation of the source text quickly in a cost-effective way with as little human intervention as possible. | To achieve human quality in a cost-effective way with a shorter turnaround time compared to human translation. | To use the available multilingual reference material or previous translations in order to achieve better quality with the same turnaround time as in the case of a general machine translation and post-editing workflow. | To achieve human quality in a cost-effective way in a shorter time compared to general machine translation and post-editing. |
Quality | The text is less natural than at the end of a purely human workflow, but fluent enough not to hinder easy comprehension. The translation is perfect content-wise with some potential minor errors that do not affect the meaning of the text. | In the case of full post-editing, the final text meets all criteria for human translation. | The translation contains meaningful target-language sentences, which are similar to the reference materials in terms of style and wording. It typically follows the terminology of the previous materials with some potential minor errors. | In the case of full post-editing, the final text meets all criteria for human translation. |
Typical content | Educational material Instruction manuals Technical documentation Texts for internal use Internal communication messages, Internet comments, Private mail, chats, forums, Customer service correspondence, Regulations |
Websites, Blog posts, Frequently asked questions, Social media posts |
All content that is similar to the reference corpus |