Learn what to do to avoid mistakes when using this type of technology
Among so many tools, applications, software and services developed with the aid of technology, machine translation is an attractive solution because it allows people who can’t speak a certain language to be in contact with its contents.
However, while bringing so many benefits, its use should be done with moderation and caution, since you are very likely to come across a translation that’s very different from the one originally desired, and this can be harmful depending on the application.
To help you out, we have selected some tips to avoid these issues and improve your experience with these software solutions. But first let’s look at how significant their market share is.
Is machine translation widely used?
Yes, absolutely. As we have talked about it in other posts, Google Translate alone, the leading translation app, translates at least 100 billion words a day, according to a Google release in 2015. That’s 4.16 billion words per hour, 69.44 million per minute, 1.15 million per second and 1,157 per thousandth of a second.
In addition, the high amount of money machine translation turns around worldwide is also noteworthy, not only Google, but the service as a whole. Check out some statistical information and their respective sources:
The machine translation market was valued at USD 122.3 million in 2016 and is estimated to grow with a Compound Annual Growth Rate (CAGR) of 6.7% from 2017 to 2023 (Prescient & Strategic Intelligence).
The machine translation market is estimated to reach USD 983.3 million by 2022, with a CAGR of 14.6% between 2012 and 2022 (Grand View Research).
Between 2017 and 2024, the CAGR of the machine translation market is estimated to reach 19%, which would make the market exceed the USD 1.5 billion mark reached in the last year mentioned (Market Study Report).
Figures vary according to each study because they were performed on different dates and with different methods, techniques and estimates, but the power of machine translation is clear, a tool that gets stronger every day and that should stay that way for a long time.
Useful tips on avoiding problems with machine translation
Despite significant growth and the use of technologies such as machine learning to provide better translations, machine translation can still cause you some problems.
Therefore, learn how to prevent them:
Beware of literality
There are some dangers in relying on machine translation, such as literal translations. While there is a high investment in machine learning and user experience optimization, literal translations still occur from time to time.
Some words don’t have only one meaning and their meaning can be fully related to the context – these words are called homographs or polysemic words, that is, words that can be used in different cases and for totally different situations.
“Letter”, from English, for example, can mean both “letra” and “carta” in Portuguese, depending on the context used, as well as the word “match”, which can be “fósforo”, “partida” (like in soccer, basketball etc.) or the verb “combinar”.
Machine translation is not always able to recognize these variations and may yield a translation result that is far from what the original sentence really meant.
Go to Wikipedia
As for the previous tip, a suggestion can help you know what an expression means, although this is usually limited to definitions and concepts, without the possibility of knowing the meaning of whole sentences, for example.
Search Wikipedia for the word or phrase in the source language. After that, in the menu, look for that article in another language – the menu appears on the left side of your computer screen and in a button with the translation symbol on your phone or tablet.
Words and phrases tend to be contextualized in Wikipedia, i.e. the meaning is adapted to each language, different from what machine translation usually does. This website can be useful if you are looking for more assertive answers.
Do reverse translations
Another handy tip on machine translation is reverse translation. Start translating from the source to the target language, and then translate the result back to the source language.
This can help you identify homograph words if the translated term results in a different word when returned to the source language. In this case, caution is necessary when translating this specific word.
Check sentence or term definition
Machine translation apps often offer a translation and a brief explanation of what the term or expression means in its original language, as in a dictionary.
To avoid being caught off guard by polysemic words, it is worth translating the meaning as well. Then, you will be able to think, in your native language, of a word that can be applied to define what you want, even without the service being done directly by machine translation.
Do not translate important documents and content
Last but not least, we should mention that machine translation really helps in infrequent and personal situations, but when it comes to something professional and of greater relevance and importance, this option should be discarded.
We have previously seen that the market value of machine translation is quite high. Also, anyone who uses these applications or websites may find that quality has been improving over time, but it still doesn’t compare to the quality of a professional translator service.
Whenever you need a professional service, look for a specialized agency that can meet your needs with excellence. You will avoid problems, and serious financial and credibility losses.
Machine translation: convenience has its price
It is undeniable that machine translation helps us on certain occasions, but it is impossible to say that it can replace professional translators working in the market. The theoretical and practical knowledge these professionals have needs to be applied to each service offered.
When you need a professional service, count on Fidelity, a translation agency with over 49 years of market experience and a highly qualified team working with excellence. This will help you avoid the pitfalls of machine translation.