The technology behind machine translation is being constantly developed and improved. Nowadays many people have translated something automatically. The best things about machine translators are wide availability and fast translation. There are many kinds of translators available in the Internet. Some of those can be purchased and some are totally free. Likewise those tools differ in supported file formats and technical features. Without too much generalization, it can be said that in some cases the price indicates the versatility of the tool.
One of the latest machine translator innovations are translator gadgets, which are desktop gadgets for PC desktops. Translator gadgets are usually very simple tools to use: After you have downloaded the gadget you can start using it. First select the language of the original language and the target language. Then just start typing or just copy and paste pieces of text from a document or from the Internet. These gadgets will automatically write the text in the other language for you.
This technology can be very useful because machine translator produces quality which is of “good enough”. However it is reasonable to highlight that the quality of automatically produced text cannot yet reach the quality level of a professional human translator. As mentioned, the risk in using machine translation is commonly related to the quality. Most of the time the quality is good but sometimes machine translators can make mistakes when choosing between words that seems to be synonyms. It is mission impossible to know which translations are mistaken if you don’t speak that language yourself.
Today’s technology enables new innovations in the area of machine translation quality. While more and more information about languages are collected to translation memories, the quality itself is improving. However there is still a lot to do if the target is the quality of a professional human translation. Another way to improve the quality is to perform automatic translation quality checking. This means that one will receive an estimation of how good the translation is. Technically speaking these estimations are usually based on some kind of comparison and match calculation because that is something a computer can do easily.
The way how the quality estimation is presented to the user can vary a lot. For example colour codes are one way to illustrate quality. Green color means that the translation is probably good, yellow will instruct you to be cautious and red color indicates that you should have a professional human translator to check the translation before using it in public. Also percentages or numbers or other coding methods can be used. No matter how the estimation is presented, it is still important to keep in mind that estimation is based on technology which is possible for a computer; it is an estimation and not the ultimate truth. Even people can have different opinions about the writing and spelling rules!
The key in machine translation is to know whether you can trust that the translation is good or not. This has been a problem which has led to the slightly negative reputation over the automatic, computer aided translators. There have been too many examples of automatically translated texts which have been used publicly without any proofreading. The fear of making mistakes has impacted the public attitude towards this technology. Automatic quality estimation is still a new thing but already it is trying to fade this fear away by giving an estimation of the translation quality.