Machine translation (MT) is a process where computer software is used to translate text from one language into another. Global Lingo utilises MT technologies to improve our translator productivity. Which in turn, offers our clients the high-quality translation they’ve come to expect with Global Lingo. With even faster turnaround and cost savings. We use a mixture of machine and human translators in our workflow process. This is to create accurate translations, in record speed.
A short history of machine translation
The first machine translation was created in the early 1950s. It was a very simple system that could only translate between Russian and English. This project was known as “The Georgetown-IBM Experiment”.
In the 1970s, MT began to be used commercially. SYSTRAN was one of the first companies to sell machine translation software. However, it was not very accurate and often produced nonsensical results.
In the 1980s, there was a renewed interest in MT. This was with the development of statistical and rule-based methods such as the IBM Candide system. When the 1990s came along, there was a huge increase in interest in MT. Many solutions became more accessible to small businesses with the development of personal computer based systems.
Today, MT is used by many large companies such as Google and Microsoft. More accessible solutions are vastly offered to small businesses and individuals to translate text quickly and easily.
What are the benefits of using machine translation?
MT can be used as a first step in the translation process in order to create a draft translation. This can then be edited and improved by a human translator. MT is also beneficial for large projects with tight deadlines, as it can speed up the process. We can pinpoint 3 main reasons to incorporate machine translations into your workflows:
- It can be used as a first step in the translation process. To create a draft translation, allowing for human translators to efficiently add value to tedious work now completed by machines.
- It is beneficial for large projects with tight deadlines, as it offers considerable time and process efficiencies.
- Machine translation can improve the consistency of translations. MT can use trained algorithms that match specific terms and phrases to their relevant context.
When to use machine translation
The machine translation process can be confusing if you don’t know what to expect. Here we’ll explore different examples and see which workflows will produce the best result for your needs.
Free and easily accessible online machine translation
Machine translation is often used as an example in the context of a “free” machine based solution. Utilising a generic machine translation system, such as Google Translate. Is fine for ad hoc translations of short and simple messaging.
This is where quality and data security are not of importance. However, this type of approach is not suitable in the workplace. It can often lead to security concerns and lack of control over content being translated.
Machine translation on its own
Partnering with an LSP for machine translation is great if you want a quick result. While retaining control over content and security, but lack the time or expertise to do it yourself. The engine will automatically translate text for use in whatever way necessary without any human intervention.
Machine translation is often appropriate for databases, webchats, and user generated content. This can include crowd-sourced Q&As, as well as other items where exact translations aren’t necessary or feasible. However, these translations may not be 100% accurate. We’d only recommend this type of usage for informal and internal communications.
Machine translation + post editing
This is the most accurate and reliable type of workflow as it involves two steps: machine translation and post-editing. A machine translates the text. A human editor will then revise it to improve accuracy, style, and readability, and a final check for any errors.
This workflow is best suited for important documents such as legal contracts, marketing materials, and website content. At Global Lingo, we pride ourselves on the value we add to machine translations through our post editing process.
Now that we’ve gone over the different types of machine translation workflows. Let’s take a look at some examples of when MT can be used.
Machine translation can be used for:
- Short and simple messaging
- Large projects with tight deadlines
- Improving the consistency of translations
When choosing a machine translation workflow, keep in mind your needs and objectives. If you want a quick result without compromising quality, we recommend using a machine translation + post-editing workflow. This will ensure that your text is accurate and error-free.
What are the challenges, and ‘best-practice solutions’ to using machine translation?
MT can often create inaccurate translations. This is because it relies on algorithms rather than human intelligence to translate the text. Furthermore, MT solutions on their own can’t match the human understanding of nuance, tone, and delivery. It is important to have a human translator check the accuracy of the translation before it is published.
Our philosophy at Global Lingo is to understand and appreciate the advantages of Machine Translation technologies. We incorporate them into the wide range of language service solutions offered to our clients.
We approach this challenge with our human post editing services. These services offer multi-step review of the translations provided, ensuring they are refined and communicate exactly how you intend.
Getting started with machine translation
When it comes to getting started with MT, Global Lingo can help. We offer a range of language services that utilise MT technologies. So whether you’re looking to streamline your workflows or improve your communications, we can assist. Our team of experts will work with you to assess your specific needs and tailor a solution that meets them.