There’s a huge need for high-speed, quality translation but at affordable rates and with human verification touchpoints. Enter post-edited machine translation.
Machine translation is an accurate and workable solution, but it has yet to reach its pinnacle.
The use of computer technology to translate a piece of text into a different language with zero, or minimal, human involvement speeds things up and has proven to be effective.
You can even translate using ChatGPT these days.
Still, human intervention remains key, especially in the verification stages.
Yes, accuracy is important, and machines can perform word for word translations but about voice and tone? What about context and cultural considerations?
That’s where post editing comes in, and you can make use of varying degrees of human touchpoints.
What is post-edited machine translation?
Machine translation engines can translate thousands of words in minutes, whereas a human can translate 3000–3500 words per day. Quicker turnaround times are a hallmark of MT, not to mention the cost savings.
You can, however, still get the assurances that come with human quality checks in the post-editing process. The costs will escalate, but the output quality improves as well.
Post-editing therefore refers to the revising and refining of content after it has been machine-translated.
A human translator checks the content to make sure it’s technically accurate, has the right voice and tone, and is culturally appropriate.
With the help of AI, it becomes infinitely easier for brands to localize their content for target audiences, for example. But we still need human review for more in-depth localization and translation projects for the highest levels of accuracy.
Think of it as an optimization of the work done by the machine learning model. The level to which we post-edit depends entirely on the needs of the client.
Types of Post-Edited Machine Translation
There are basically two types of post editing: light and full.
Light post-editing involves as few modifications as possible. Its goal is to make the translation understandable, factually accurate, and grammatically correct.
Think of it like the difference between a continental breakfast and a full brunch.
One’s complimentary or at least light on the wallet and will keep you going for a little bit with the essential nutrients. The other offers more options and will fill you up for a while, but comes at a greater cost.
One isn’t better than the other, per se, but often comes down to what you need.
Linguists perform as few modifications as possible to the raw machine translation output to make the translation understandable, factually accurate and grammatically correct.
This includes smaller tasks like:
- correcting grammar, punctuation, and spelling mistakes
- rewriting confusing sentences, but only if they clarify the meaning of the text.
So, the light post-editing approach only covers critical errors and or poor use of terminology that can cause unintelligible content or unintended legal consequences.
It might not flow, but the text is good enough and, most importantly, factually correct.
This is a slower and more detail-oriented process. The goal remains producing accurate translations with consistent and correct terminology, appropriate tone, and style. There will also be no stylistic inconsistencies or grammatical mistakes.
So, after the full post-editing step, the text should look as if a professional linguist worked with it from the very beginning.
Full post-editing tasks include all the light post-editing tasks, plus:
- cross-referencing translations against other resources
- making syntactic modifications to match the target language.
The content must therefore meet the precise quality criteria defined by the client for human translations.
Benefits of Post-Edited Machine Translation
Human translation is slow and sophisticated, and it comes at a high cost, but the quality is genuine and premium.
Machine translation is quick and simple, and you can get it done cheaply. The quality, though, isn’t great.
And by the way, it’s also not as simple as choosing between light and full post-editing. It all differs from project to project, client to client, language to language.
But with post-edited machine translation, you can achieve some semblance of balance.
We put this simple chart together to help you get thinking about your translation priorities.
If your text requires perfection, a professional human translator or editor will need to work on your content. You get the highest quality, but it comes at a cost.
If speed of execution is the main factor, MT must be part of the solution. It gets the work done at a much lower rate, but you sacrifice quality in the end.
Machine translation with human quality assurance is where you’re likely to find the balance between quality and cost efficiency.
It’s the PEMT sweet spot: A combination of MT plus post-editing and AI evaluation can help achieve lower price points without sacrificing quality or throughput time.
Let PEMT Work for You
There’s no one way to approach translation projects. Indeed, we’re quite firm on flexibility around here, and offer the specialized services to back it up. What you need is what you’ll get one way or another, and anywhere in between.
And that’s the thing. It’s a sliding scale between pure human translation and handing everything over to the machine.
At Summa Linguae, we consider all the factors before rolling out a custom price.
First and foremost, you must know what exactly you need. Make a list of the languages into which or from which you need to translate and decide on the level of language services your company needs.
Will you need translation or localization? And is it a one-time need, or will there be ongoing projects?
The more detailed information you provide to the person quoting for your translation needs, the greater the chance that you will be offered the most attractive price. And if you don’t know exactly what you need, we can help figure it out.
Contact us today to get started.
Pending No More: Move Ahead with Patent Translation Services
Patent translation services require a balance of language and technical expertise, and a flare for legalese.
Quality Insurance: Leveraging AI for the Best Machine Translation
Run it through the machine translation process, test accuracy with AI, then comb it over with post-editing...