The Influence of Machine Translation for Asian Languages

Nowadays, people find their lives have been significantly changed by the Internet, and languages are no longer the huge barrier for reading and searching on the Internet, thanks to machine translation tools that have become their useful and powerful assistants.

Machine Translation (MT) has been widely used in many areas and industries besides manual translations. Some government departments and huge organizations in the US have deployed Machine Translation solutions for several years to translate their materials into other languages, and these kinds of MT tools have significantly sped up their translation and reduced their costs. As for quality, these MT tools have ofen produced satisfactory results for European languages (especially French, Italian, Spanish and German, the so-called ‘FIGS’ languages) when the source language is English. But things are different when the source language is English and the target languages are double-byte Asian languages like Chinese, Japanese and Korean.

Due to many factors, including the impact of cultural context, very different grammar, differing styles of rhetoric, etc., Machine Translation has not achieved the same results as for European languages. This is mainly caused by the language properties. European languages  are based on alphabets in which letters represent sounds, combined in different ways to represent words/ideas (but with the individual letters themselves having no meaning). But East Asian languages (such as Chinese, Japanese and Korean) consist of characters representing concepts, not sounds, where each character is different and two or more characters can make different words. These factors combine to make things more complex for machine translation.

These challenges notwithstanding, though, machine translation has made great achievement in double-byte languages in recent years, especially in Chinese. Until relatively recently, MT often yielded poor results, outputting eccentric and tricky sentences,, and often just plain garbage. But now many web users can enjoy the benefit of MT by using it when searching the Internet. For these users, knowing the outline and the main idea of the webpage is enough: there’s no need to get very accurate and good expressions of the content. For example, Microsoft MSDN and Technet articles have been fully translated using Machine Translation tool, and the final output is really good for those engineers who want to solve their technical problems.

But for translations requiring highly accurate rendering, old-fashioned human translation is still required in most cases when going from English into Asian languages. As a result, most business content still uses human translation. However, it needn’t be all one or the other: owevermany companies still realize the cost and scalability potential of MT by combining it with human review/editing at varying levels to achieve the required quality. For example, a document requiring a level of accuracy greater than that achieved by raw MT, can be subjected to post-editing to ensure it is brought up to the right level of quality. But you need to make sure you are partnered with experts like EC Innovations to help ensure that you are striking the right balance and making the right decisions on when to use v not use MT. Otherwise, you can end up spending more money on MT + Post-editing than you would have spent on human translation….and still not have the needed quality!

All in all, machine translation can help to control translation costs, shorten lead times, and improve the efficiency of translation. We can even foresee the day when machine translation will be much more popular, even for Asian languages, and will be deployed in many areas. But human translation will still be necessary in many cases and cannot be fully replaced. So make sure you partner with the right localization provider to help you strike the right balance!


Leave a Reply