Facebook claims to have found quicker way to translate through AI

Facebook claims to have found quicker way to translate through AI
The Siliconreview
11 May, 2017

The popular social networking giant in a move to conquer language fences has recently announced all new and innovative machine learning translation method. The company claims it to nine times faster than other competitors.

According to a report by Verge published recently, Artificial intelligence (AI) has been previously in place at Facebook for translating the status updates automatically to other languages, but the company is creating a changeover from lab to app.

"We're currently talking with a product team to make this work in a Facebook environment. There are differences when moving from academic data to real environments in terms of language. The academic data is news-type data; while conversation on Facebook is much more colloquial," the report quoted Facebook's AI engineer David Grangier as saying.

The all new machine learning version method will take certain time to get executed and live as a research as of now. But Facebook has said that it will probably happen further down the line.

"Usually, AI-powered translation relies on what are called recurrent neural networks (RNNs), whereas this new research leverages convolutional neural networks (CNNs) instead," Facebook's AI engineers explained.

The RNNs analyzed date in sequence by working on it left to right from side to side a sentence in order to sort and translate it word by word while CNNs look at disparity feature of data at the same time - a style of calculation that is much improved and faster.

"So translating with CNNs means tackling the problem more holistically and examining the higher-level structure of sentences. The [CNNs] build a logical structure, a bit like linguistics, on top of the text," said Michael Auli, another Facebook AI engineer.

While the company also noted that the AI communities were eager to get better upon the commonly used RNNs for translation - a method that has get through marvelous efforts already.

"The short answer is that people just hadn't invested as much time in this, and we came up with some new developments that made it work better," Grangier added.