Home / / Microsoft creates AI translation system as good as human

Microsoft creates AI translation system as good as human

Research team achieves parity with human translators for Chinese-to-English news translation

Xuedong Huang, technical Fellow in charge of Microsoft's speech, natural language and machine translation efforts.
Xuedong Huang, technical Fellow in charge of Microsoft's speech, natural language and machine translation efforts.

A team of researchers from Microsoft say they have created an AI translation system which is as good as human translator.

The team has developed a Chinese-to-English translator, which utilises machine learning in natural language processing, which can translate documents with parity to human efforts.

The results have been verified by external bilingual human evaluators who were hired to assess the quality of translation, Microsoft said in a blog post.

Microsoft said that expects the techniques used can be applied in its translation products for other languages, including complex languages.

The system uses a mixture of existing and new techniques to train the system, in a method called deep neural networks, which creates more fluent and natural sounding translations, Microsoft said.

The blog post explained:

One method they used is dual learning. Think of this as a way of fact-checking the system's work: Every time they sent a sentence through the system to be translated from Chinese to English, the research team also translated it back from English to Chinese. That's similar to what people might do to make sure that their automated translations were accurate, and it allowed the system to refine and learn from its own mistakes. Dual learning, which was developed by the Microsoft research team, also can be used to improve results in other AI tasks.

Another method, called deliberation networks, is similar to how people edit and revise their own writing by going through it again and again. The researchers taught the system to repeat the process of translating the same sentence over and over, gradually refining and improving the response.

The researchers also developed two new techniques to improve the accuracy of their translations, Ming Zhou, assistant managing director of Microsoft Research Asia, said.

One technique, called joint training, was used to iteratively boost the English-to-Chinese and Chinese-to-English translation systems. With this method, the English-to-Chinese translation system translates new English sentences into Chinese in order to obtain new sentence pairs. Those are then used to augment the training dataset that is going in the opposite direction, from Chinese to English. The same procedure is then applied in the other direction. As they converge, the performance of both systems improves.

Another technique is called agreement regularization. With this method, the translation can be generated by having the system read from left to right or from right to left. If these two translation techniques generate the same translation, the result is considered more trustworthy than if they don't get the same results. The method is used to encourage the systems to generate a consensus translation.

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 last year. The tests included around 2,000 sentences from Chinese news reports.

The Microsoft team, which included researchers from China and the US, said they were surprised to have been able to achieve the results so quickly.

Xuedong Huang, a technical fellow in charge of Microsoft's speech, natural language and machine translation efforts, commented: "Hitting human parity in a machine translation task is a dream that all of us have had," Huang said. "We just didn't realize we'd be able to hit it so soon."

"The pursuit of removing language barriers to help people communicate better is fantastic," he said. "It's very, very rewarding."

Arul Menezes, partner research manager of Microsoft's machine translation team, said the team set out to prove that its systems could perform about as well as a person when it used a language pair - Chinese and English - for which there is a lot of data, on a test set that includes the more commonplace vocabulary of general interest news stories.

"Given the best-case situation as far as data and availability of resources goes, we wanted to find out if we could actually match the performance of a professional human translator," said Menezes, who helped lead the project.

Menezes said the research team can apply the technical breakthroughs they made for this achievement to Microsoft's commercially available translation products in multiple languages. That will pave the way for more accurate and natural-sounding translations across other languages and for texts with more complex or niche vocabulary.