Artificial intelligence (AI) is the future of global content strategies. In today’s fast-paced business environment, organisations must handle vast amounts of content across diverse platforms and channels.
What is machine translation?
Neural machine translation (NMT) relies on an algorithm that derives knowledge from human-made translations. Machine translation operates by utilising computer software that incorporates various algorithms, such as neural networks. This software examines and makes sense of the content, subsequently producing a translated version.
What types of content can be translated using machine translation?
Machine translation is used in a range of industries. It excels with structured and technical content but can also be applied to creative material. Industries such as health and life sciences, legal and financial sectors, politics, human rights, energy, environment, climate, sustainability, education and manufacturing all benefit from machine translation.
What are the benefits of machine translation?
On average, machine translation can lower costs by 30% compared to traditional, human-only workflows, while also reducing turnaround times by up to 50%.
However, the specific savings achieved may vary depending on factors such as the language pair, content type and subject matter.
Why do you need a translation partner for machine translations?
Using a machine translation engine is simple and the outcome of the process is sufficient if you only want to get the gist of a text. However, a raw machine-translated document may contain glaring grammatical errors, typos and punctuation mistakes, not to mention mistranslations, making it unsuitable for business use. That is where an experienced native speaker human post-editor comes in. Depending on the intended use, a post-editor can correct these errors and another native speaker proofreader can improve the style and structure, so that you receive a translation that is ready for publication.
Our AI and machine translation services
Our neural machine translation engines support over 40 language pairs. However, not all languages are equally well-suited to machine translation due to factors such as grammatical complexity and the availability of high-quality language data. Therefore, the potential savings achievable with machine translation vary across languages.
Our AI and Machine Translation services
Option 1:
Machine translation with light post-editing
This service includes:
- AI machine translation
- post-editing by a native-speaking translator
During light post-editing, a linguist corrects grammar and spelling mistakes, punctuation and terminology errors in the machine translation. Light post-editing does not include style, fluency and consistency improvements. The quality of the machine translation with light post-editing is not equal to a human translation. While the end result might sound less natural, it is a quicker solution and fits a smaller budget.
Following the translation process, our QA team runs a basic in-house quality check, which consists of:
– spot checking the translation to make sure there are no omissions, typos or spelling mistakes
– formatting the translation(s) to match the original source document(s).
Option 2:
Machine translation with full post-editing
This service includes:
- AI machine translation
- post-editing by a native-speaking translator
- proofreading by a second native-speaking translator
During full post-editing, a linguist corrects grammar and spelling mistakes, punctuation and terminology errors in the machine translation, and then a second linguist (proofreader) improves style, fluency and consistency.
Following the post-editing process, our QA team carries out a two-step, in-house quality check, which consists of:
– a terminology review and double-checking of names and organisations
– formatting the translation(s) to match the original source document(s)
– content, consistency and coherence check.
Please note that the outcome of the machine translation with full post-editing is different to that of a human translation. A native-speaking human translator may use a different sentence structure or style.
Please do not hesitate to contact us if you need further information.