Microsoft's AI system can translate Chinese to English with human accuracy

Microsoft Achieves Breakthrough In Chinese To English Machine Translation

Microsoft researchers match human levels in translating news from Chinese to English

Machine translation has come a long way and while it works to the extent that it enables you to understand what's written in a foreign language, there's still a lot of room for improvement to bring up the translation quality up to human standard.

Microsoft has reached a new milestone with its translation tech.

The encoder-decoder framework has achieved promising progress for many sequence generation tasks, including machine translation, text summarization, Q\&A, dialog system, image captioning, {\em etc}.

To ensure the results were both accurate and at par with what people would have done, the team hired external bilingual human evaluators who compared Microsoft's results to two independently produced human reference translations.

Microsoft Research has aspirations for the potential of this new machine translation system that involves improving the field of Artificial Intelligence research, but before soaring to great fanciful AI heights, the MTS will need to be first tested in real-world scenarios.

"Hitting human parity in a machine translation task is a dream that all of us have had", Huang Xuedong, in charge of Microsoft's speech, natural language and machine translation efforts, was quoted as saying. With this method, the English-to-Chinese translation system translates new English sentences into Chinese in order to obtain new sentence pairs.

Ex-spy poisoning: USA holds Russian Federation responsible
British Foreign Secretary Boris Johnson said Britain's quarrel was not with the Russian people but with the Kremlin. Alexander Yakovenko said threats don't come often, but the embassy takes each of these incidents in earnest.

The researchers taught the system to repeat the process of translating the same sentence over and over, gradually refining and improving the response, Microsoft said.

The team used a dual-learning method.

The successful translation used a set of news stories which were set as a standardised test of machine learning translation capabilities at a conference in September past year.

"People can use different words to express the exact same thing, but you cannot necessarily say which one is better". "We just did not realize we would be able to hit it so soon".

Though limited to a single test, the results are impressive given the difficulty technology has traditionally experienced in translating Chinese into English and vice versa. "It's very, very rewarding". Menezes said the research team can apply the technical discoveries they made in this test to Microsoft's commercially available translation products in multiple languages.

In most cases, for a news article to be accurately translated, a human translator must do it.

Latest News