Patent infringement for Korean to English translation
Will machine translation ever substitute human translators? That is not likely, since the ability to detect context and compose stylish sentences is given only to human.
In November, 2006 Supreme Court of Korea affirmed patent infringement by Microsoft of Korean patents for automatic Korean to English translation, used in the local version of Microsoft Office.
Hankuk Aviation University Professor Lee Keung-hae received patents covering automatic language translation from filings dating to 1997. A Korean company, P&IB, purchased the patent rights in 1998.
In 2000, P&IB filed an infringement suit against Microsoft Korea. Microsoft replied with invalidity assertion.
The 70 billion won ($75 million) damage award was what P&IB had sought.
Microsoft vows to fight on. Microsoft lawyer Chung Jae-hoon defiantly tooted: “We recently found critical evidence that will disprove the effects of the language-switch technology patents and we will file another lawsuit to ask for nullification of the patents.”
Microsoft faces continuing damages if it does not cease infringement, and, worst case, may have to stop selling Office in Korea if it does not come to terms with P&IB.
In May, Microsoft lost an antitrust suit in Korea, and was forced to sell a stripped-down version of Windows that did not include instant messaging and a media player. Microsoft has a similar situation in Europe, which ruled in March 2004 that Windows not include a media player. Microsoft has been fined $357 million for its refusal to comply, and faces fines of up to $3 million per day if it continues ignoring the order.
By the way, the quality of machine translation is so poor in Korean-English, it will be nice to develop it further to achieve at least 80% accuracy. I have used machine translation before and it is totally useless. If there is any software engineer who can work with me, I think I can develop a software that can achieve that much accuracy, from my 28 years of translating experience.