AAC Global Blog

Free content to inspire and educate you from our language and learning experts

Nov
21

What are the best uses for machine translation?

Companies have a broad and diverse range of translation needs. For some texts, translation quality is the highest priority. Then there are texts for which quick delivery is the most important thing. The third factor is price: there may be massive amounts of documentation stored in archives that remain untranslated simply because of the cost. For the latter two cases, there is a solution that makes everyday information sharing and communication easier, even when there is a very large amount of text.

In this blog post, we will explain which text types are best suited for machine translation.

What does machine translation mean?

In machine translation, texts are translated automatically with computer software so that a human translator does not directly participate in the process. The relationship between human translators and machine translation systems has been that the machine translation engine has been trained with translations created by human translators. The undisputed strength of machine translation is the speed of the translation process. Customers can get the translations they need in minutes.

The strength of human translators is still the quality of the translation. Machine translation engines can produce the same quality as a human translator for a few text types only. Today, machine translation works best in scenarios where a text needs to be conveyed in an understandable form in another language. The language of the translation is not necessarily error-free or fully logical, but it can be understood. The word order may be odd, or a single word may have been replaced with a strange synonym. For instance, the translation of the word ‘volume’ may refer to sound volume, a quantity of space or an amount, regardless of the actual context. However, a reader with basic language skills can understand what the intended meaning of the text is.

2019-11-06-Machine-Translation-illustration

Data security often comes up when discussing machine translation. A machine translation engine dedicated to the company’s use is a safe option in terms of data security. In practice, the material to be translated is transmitted in an encrypted form between the client and the machine translation system. The translated material only remains in the memory of the machine translation engine for a short time, and it is not used for any other purposes. This way, even texts containing business secrets can be translated using machine translation. With free, publicly available machine translation services, there is a risk that the translated text will remain in the translation engine’s memory. Often, the party offering a free translation service withholds the right to use the translated material.

At present, machine translation can well be used, for instance, for non-public, company-internal communication and information sharing, especially when the content of the text needs to be transmitted quickly but the text is not distributed to a large audience. It is also important that all parties are aware of the use of machine translation.

What are the best uses for machine translation?

Machine translation can be used in a large variety of situations. In the following, we will go through some scenarios where machine translation can help with a company’s daily operations.

Internal emails and other written communication

If a company is operating in several countries or on several continents even, communication becomes more difficult to manage. The language skills of employees can vary a lot. Even if English was chosen as the company’s official language, not everyone may be able to use it, at least well enough. Many people would in fact prefer communication in their native language, which would decrease ambiguity. However, this is nearly impossible when the company has sites all over the world.

Machine translation can help decrease or even eliminate the language barrier in communication. A translation of the text can be obtained quickly and in a form that enables the recipients to understand the core content of the message.

Machine translation can be particularly important in cases where an acute issue needs to be communicated to a large number of people. When, for instance, a disturbance in production can be announced quickly to a large group of people, the problem can also be responded to quickly and effectively. In such a situation, the recipient can quickly and easily translate the original message into their own language, or the sender can take care of that.

Even though machine translation solutions are developing all the time, it must be kept in mind that the translations they produce are not completely free of errors. So one cannot assume at this point that the output of a machine translation engine is always good.

Machine translation can be used in many ways, for instance, to follow the development of one’s field of expertise and to obtain competition-related information. The bulletins and advertisements of a local competitor can be translated, as well as the content of local patents.

Machine translation can help in situations where speed is the most important factor. You get the machine translation almost immediately for your use. It is also a great benefit that machine translation can be used to easily translate texts that would otherwise not be translated at all, such as email messages.

Documentation and instructions

Companies operating in different industries can accumulate considerable amounts of documentation on products, services and operating models. Often, all relevant information is recorded for future reference, either for development actions or just for maintenance. As is well known, shared information is the best kind of information, but in a multinational company, the documentation would first need to be translated into several languages. The resulting cost is often considered too high.

There may also be large quantities of various instructions, but translating them is not considered a viable option due to the cost.

Machine translation can also be used to translate PowerPoint presentations, intranet bulletins, and other similar documents. The need to translate them is often urgent.

Machine translation is an excellent choice in situations where having the materials translated by a human translator would be too large an investment.

The goal is to share information within the entire organization so that as many employees as possible have access to the best information that supports for their work. Machine translation helps the organization share essential information better.

If necessary, post-editing can be ordered for selected important texts, so that a human translator edits the machine translation into a more refined form.

One benefit of AAC’s machine translation service is that a file can be attached to the machine translation order; there is no need to cut and paste the text. For urgent situations, an immediate translation can be ordered through the portal. This option is for situations where a translation is needed as soon as possible.

2019-11-06-Machine-Translation-illustration2

Texts with a short life cycle

A good example of texts with a short life cycle is that of product review texts at online stores. From them, customers may get information crucial to their purchase decision that cannot be found elsewhere. A written review provides valuable information both to the company and customers visiting the company's website.

When an online store is operating in multiple countries, using machine translation for translating reviews can be an excellent way to serve customers. Product reviews can be easily translated into another language. Customers of the online store can read the review in their own language. Here, machine translation helps customers with their purchase decisions. Usually, the most important thing for the reader is to understand the essential content of the review; it does not matter if the machine translation engine occasionally makes some silly mistakes or formulates sentences in a clumsy way. Machine translation can be integrated into the translation process so that the translations can be published automatically, or quality assurance by human professionals can be connected to the process.

Machine translation systems integrated into an online store are often used in the travel industry’s online services and the online trade of consumer products. The reviews are typically translated into the language of the customer's web browser.

When the product catalog of an online store is extensive, machine translation is quite commonly used for translating product information. In these cases, it is a good idea to classify products into more and less important ones, so that human quality assurance can be added for the translations of the most important products, if necessary. For less important products, the quality of machine translation may well be sufficient, for instance, if the product description only includes a couple of sentences.

Integrated machine translation systems are commonly used in social media services where the posts of users can be translated into the reader’s own language with a single click.

When better quality is needed

The translation quality produced by machine translation is not sufficient as such in many situations. If the documentation to be translated needs to be easy to understand and, especially, if the documentation is critical in terms of safety, the appropriate solution is post-edited machine translation or traditional translation by a professional translator. In the translation process with post-editing, the source material is first translated into the target language with machine translation, and a human editor then reviews and corrects the translation.

The idea in post-edited machine translation is that a human corrects any obvious errors made by a machine. In addition, the end result of post-editing can be used to train the machine translation engine to further improve its operation. One of the benefits of this process is the delivery time – it is often shorter than with translation performed by a human translation from start to finish. However, the shorter schedule requires that the original machine translation output is of sufficiently good quality.

A good example of training a machine translation engine is a situation where a repeating error occurs in the translations produced by the engine, such as the English names of weekdays written in lower case. Such systematic errors can easily be fixed by training the translation engine further.

For the translation of user manuals, marketing materials and other public documents, a human translator is still recommended. A human professional can introduce deeper meanings and desired nuances into the text. Only a human can interpret, for instance, the multi-level meanings of a journalistic or literary text. Machine translation is not fully replacing human translators, but it can help humans with texts that would otherwise probably not be translated at all.

Machine translation helps with information sharing and communication

Today’s machine translation systems can help with companies’ internal and, in some cases, external communication – at least when the text they produce is reviewed before publication. Machine translation is also useful in the translation of internal documentation. There is no need for the produced content to be available in only one language when the material can be effortlessly translated into other languages. This way, information can be distributed to a considerably wider target audience within the company. Information is most useful when it is shared within the whole organization.

However, machine translation is not a solution for all translation needs. When an understandable, useful and reviewed text is needed fast, the combination of machine translation and post-editing can be the right choice. When the best translation quality is called for, or the text type is difficult for a machine translation engine, the use of a human translator should be selected.

Does your company have a lot of internal documentation that needs to be translated into other languages? Do you want to make the company’s internal, international communication easier? Please contact us, and our experts will help you find the solution that best meets your needs.

 

Download our free e-book on machine translation!

Machine translation Post-editing Human translation

Antti Voutilainen

Antti Voutilainen

Antti Voutilainen works as a marketing content writer producing engaging and relevant content for our customers. His experience in customer service and sales has taught him how vital true customer insight and effective content marketing are today. Now his focus is only helping customer to find the perfect solution for their specific needs.

Get info about newest blogposts

Topics

See all