Before starting an MTPE post-editing project, it is important to know what is going to be translated, which languages are involved, the size of the project, what format the content is in, and what it is going to be used for. If there is already a machine translation, it is also essential that we know what the quality of that first version is like.
Why does all this information matter? Because not all content or all machine translations share the same conditions. At blarlo, these aspects are taken into consideration to establish the workflow and post-editing level for the project.
What type of content is going to be worked on?
The type of content is one of the first aspects that need to be identified.
A catalog with thousands of product descriptions will have different characteristics to a help center, a software interface, or a technical manual, for example. The size, repetition, terminology, and the usage of the content change from one project to another.
At blarlo, we provide MTPE machine translation with human post-editing for large-scale, recurring, technical, digital, or documentation content for ecommerce, catalogs, help centers, software, and technical documentation.
Where and who is going to use the translation?
It is also important to know what will happen with the text once it has been translated.
Is it only going to be used inside the company? Is it going to be published on a website? Will customers read it? Is it going to be part of a digital product?
This information helps to establish the level of post-editing needed. At blarlo, we offer light post-editing, aimed at comprehension, internal usage, or low-exposure content, and full post-editing, for visible or publishable content that requires a more thorough revision of the meaning, terminology, style, and fluency.
As such, explaining what the end use of the content will be right from the outset will make it possible to more accurately establish what work will be needed.
Why does the quality of the machine translation matter?
The post-editing is based on a translation that has been automatically generated. That’s why the quality of that first version also impacts the project.
An automatic output may contain mistakes in the meaning, omissions, terminological issues, or sentences that sound natural but don’t correctly convey the original meaning.
If an already exists, having a sample will enable us to see what we are starting with. At blarlo, the quality of that output is assessed together with factors such as the complexity, size, timeframe, channel, and exposure of the content.
What if the company already has pre-defined terminology?
This can also be incorporated into the project.
If there are product names, approved terms, glossaries, translation memories, or style guides, these resources help to maintain common criteria across files, languages, and updates.
This is particularly relevant in recurring projects. Consider, for example, an ecommerce that adds new products every week or a type of software that regularly updates its help center. Maintaining the same terminological decisions prevents inconsistencies as and when the content grows.
What information is it worth providing?
To define an MTPE post-editing project, the main details are:
- the source and target languages;
- the type of content and sector;
- the size and format;
- the timeframe;
- the end use and level of exposure;
- a sample of the machine translation, if applicable;
- any glossaries, memories, or terminology instructions available.
With this information, we can more accurately establish the workflow, the level of post-editing, and the controls needed for that project.
Frequently asked questions
Not necessarily. If you already have one, a sample enables us to analyze the quality of the automatic output and see the starting point for the project.
Yes. At blarlo, we can work with approved terminology, product names, glossaries, translation memories, and style criteria to maintain consistency across the content.
Because blarlo offers light and full post-editing and the level of intervention varies depending on the usage, visibility, and characteristics of the content.
Yes. The service covers management of large volumes, different languages, deliveries, priorities, and follow-up to maintain consistency across different batches and updates.
At blarlo, we offer a machine translation post-editing service for content that requires speed without sacrificing human checks, with workflows adapted to the content, the volume, and the end usage. Request an MTPE quote.
