Tag Archives: Machine Translation

Zero-Shot Translation with Google’s Multilingual Neural Machine Translation System

In the last 10 years, Google Translate has grown from supporting just a few languages to 103, translating over 140 billion words every day. To make this possible, we needed to build and maintain many different systems in order to translate between any two languages, incurring significant computational cost. With neural networks reforming many fields, we were convinced we could raise the translation quality further, but doing so would mean rethinking the technology behind Google Translate.

In September, we announced that Google Translate is switching to a new system called Google Neural Machine Translation (GNMT), an end-to-end learning framework that learns from millions of examples, and provided significant improvements in translation quality. However, while switching to GNMT improved the quality for the languages we tested it on, scaling up to all the 103 supported languages presented a significant challenge.

In “Google’s Multilingual Neural Machine Translation System: Enabling Zero-Shot Translation”, we address this challenge by extending our previous GNMT system, allowing for a single system to translate between multiple languages. Our proposed architecture requires no change in the base GNMT system, but instead uses an additional “token” at the beginning of the input sentence to specify the required target language to translate to. In addition to improving translation quality, our method also enables “Zero-Shot Translation” — translation between language pairs never seen explicitly by the system.
Here’s how it works. Let’s say we train a multilingual system with Japanese⇄English and Korean⇄English examples, shown by the solid blue lines in the animation. Our multilingual system, with the same size as a single GNMT system, shares its parameters to translate between these four different language pairs. This sharing enables the system to transfer the “translation knowledge” from one language pair to the others. This transfer learning and the need to translate between multiple languages forces the system to better use its modeling power.

This inspired us to ask the following question: Can we translate between a language pair which the system has never seen before? An example of this would be translations between Korean and Japanese where Korean⇄Japanese examples were not shown to the system. Impressively, the answer is yes — it can generate reasonable Korean⇄Japanese translations, even though it has never been taught to do so. We call this “zero-shot” translation, shown by the yellow dotted lines in the animation. To the best of our knowledge, this is the first time this type of transfer learning has worked in Machine Translation.

The success of the zero-shot translation raises another important question: Is the system learning a common representation in which sentences with the same meaning are represented in similar ways regardless of language — i.e. an “interlingua”? Using a 3-dimensional representation of internal network data, we were able to take a peek into the system as it translates a set of sentences between all possible pairs of the Japanese, Korean, and English languages.

Part (a) from the figure above shows an overall geometry of these translations. The points in this view are colored by the meaning; a sentence translated from English to Korean with the same meaning as a sentence translated from Japanese to English share the same color. From this view we can see distinct groupings of points, each with their own color. Part (b) zooms in to one of the groups, and part (c) colors by the source language. Within a single group, we see a sentence with the same meaning but from three different languages. This means the network must be encoding something about the semantics of the sentence rather than simply memorizing phrase-to-phrase translations. We interpret this as a sign of existence of an interlingua in the network.

We show many more results and analyses in our paper, and hope that its findings are not only interesting for machine learning or machine translation researchers but also to linguists and others who are interested in how multiple languages can be processed by machines using a single system.

Finally, the described Multilingual Google Neural Machine Translation system is running in production today for all Google Translate users. Multilingual systems are currently used to serve 10 of the recently launched 16 language pairs, resulting in improved quality and a simplified production architecture.

A Neural Network for Machine Translation, at Production Scale

Ten years ago, we announced the launch of Google Translate, together with the use of Phrase-Based Machine Translation as the key algorithm behind this service. Since then, rapid advances in machine intelligence have improved our speech recognition and image recognition capabilities, but improving machine translation remains a challenging goal.

Today we announce the Google Neural Machine Translation system (GNMT), which utilizes state-of-the-art training techniques to achieve the largest improvements to date for machine translation quality. Our full research results are described in a new technical report we are releasing today: “Google’s Neural Machine Translation System: Bridging the Gap between Human and Machine Translation” [1].

A few years ago we started using Recurrent Neural Networks (RNNs) to directly learn the mapping between an input sequence (e.g. a sentence in one language) to an output sequence (that same sentence in another language) [2]. Whereas Phrase-Based Machine Translation (PBMT) breaks an input sentence into words and phrases to be translated largely independently, Neural Machine Translation (NMT) considers the entire input sentence as a unit for translation.The advantage of this approach is that it requires fewer engineering design choices than previous Phrase-Based translation systems. When it first came out, NMT showed equivalent accuracy with existing Phrase-Based translation systems on modest-sized public benchmark data sets.

Since then, researchers have proposed many techniques to improve NMT, including work on handling rare words by mimicking an external alignment model [3], using attention to align input words and output words [4] and breaking words into smaller units to cope with rare words [5,6]. Despite these improvements, NMT wasn't fast or accurate enough to be used in a production system, such as Google Translate. Our new paper [1] describes how we overcame the many challenges to make NMT work on very large data sets and built a system that is sufficiently fast and accurate enough to provide better translations for Google’s users and services.
Data from side-by-side evaluations, where human raters compare the quality of translations for a given source sentence. Scores range from 0 to 6, with 0 meaning “completely nonsense translation”, and 6 meaning “perfect translation."
The following visualization shows the progression of GNMT as it translates a Chinese sentence to English. First, the network encodes the Chinese words as a list of vectors, where each vector represents the meaning of all words read so far (“Encoder”). Once the entire sentence is read, the decoder begins, generating the English sentence one word at a time (“Decoder”). To generate the translated word at each step, the decoder pays attention to a weighted distribution over the encoded Chinese vectors most relevant to generate the English word (“Attention”; the blue link transparency represents how much the decoder pays attention to an encoded word).
Using human-rated side-by-side comparison as a metric, the GNMT system produces translations that are vastly improved compared to the previous phrase-based production system. GNMT reduces translation errors by more than 55%-85% on several major language pairs measured on sampled sentences from Wikipedia and news websites with the help of bilingual human raters.
An example of a translation produced by our system for an input sentence sampled from a news site. Go here for more examples of translations for input sentences sampled randomly from news sites and books.
In addition to releasing this research paper today, we are announcing the launch of GNMT in production on a notoriously difficult language pair: Chinese to English. The Google Translate mobile and web apps are now using GNMT for 100% of machine translations from Chinese to English—about 18 million translations per day. The production deployment of GNMT was made possible by use of our publicly available machine learning toolkit TensorFlow and our Tensor Processing Units (TPUs), which provide sufficient computational power to deploy these powerful GNMT models while meeting the stringent latency requirements of the Google Translate product. Translating from Chinese to English is one of the more than 10,000 language pairs supported by Google Translate, and we will be working to roll out GNMT to many more of these over the coming months.

Machine translation is by no means solved. GNMT can still make significant errors that a human translator would never make, like dropping words and mistranslating proper names or rare terms, and translating sentences in isolation rather than considering the context of the paragraph or page. There is still a lot of work we can do to serve our users better. However, GNMT represents a significant milestone. We would like to celebrate it with the many researchers and engineers—both within Google and the wider community—who have contributed to this direction of research in the past few years.

We thank members of the Google Brain team and the Google Translate team for the help with the project. We thank Nikhil Thorat and the Big Picture team for the visualization.

[1] Google’s Neural Machine Translation System: Bridging the Gap between Human and Machine Translation, Yonghui Wu, Mike Schuster, Zhifeng Chen, Quoc V. Le, Mohammad Norouzi, Wolfgang Macherey, Maxim Krikun, Yuan Cao, Qin Gao, Klaus Macherey, Jeff Klingner, Apurva Shah, Melvin Johnson, Xiaobing Liu, Łukasz Kaiser, Stephan Gouws, Yoshikiyo Kato, Taku Kudo, Hideto Kazawa, Keith Stevens, George Kurian, Nishant Patil, Wei Wang, Cliff Young, Jason Smith, Jason Riesa, Alex Rudnick, Oriol Vinyals, Greg Corrado, Macduff Hughes, Jeffrey Dean. Technical Report, 2016.
[2] Sequence to Sequence Learning with Neural Networks, Ilya Sutskever, Oriol Vinyals, Quoc V. Le. Advances in Neural Information Processing Systems, 2014.
[3] Addressing the rare word problem in neural machine translation, Minh-Thang Luong, Ilya Sutskever, Quoc V. Le, Oriol Vinyals, and Wojciech Zaremba. Proceedings of the 53th Annual Meeting of the Association for Computational Linguistics, 2015.
[4] Neural Machine Translation by Jointly Learning to Align and Translate, Dzmitry Bahdanau, Kyunghyun Cho, Yoshua Bengio. International Conference on Learning Representations, 2015.
[5] Japanese and Korean voice search, Mike Schuster, and Kaisuke Nakajima. IEEE International Conference on Acoustics, Speech and Signal Processing, 2012.
[6] Neural Machine Translation of Rare Words with Subword Units, Rico Sennrich, Barry Haddow, Alexandra Birch. Proceedings of the 54th Annual Meeting of the Association for Computational Linguistics, 2016.

Chat Smarter with Allo

At Google, we are continuously building products powered by Machine Learning to delight our users and simplify their lives. Today, we are excited to talk about the technology behind Allo, a new smart messaging app that uses the power of neural networks and Google Search to make your text conversations easier and more productive.

Just like Smart Reply for Inbox, Allo understands the conversation history to generate a set of suggestions that the user will likely want to respond with. In addition to understanding the context of your conversation, Allo learns your individual style, so the responses are personalized for you.
How does it work?

About a year ago, we started exploring how we can make communication easier and more fun to use. The idea of Smart Reply for Allo came from my teammate Sushant Prakash who, along with Ori Gershony, led their teams to build this technology. We began by experimenting with neural network based model architectures which had proven to be successful for sequence prediction, including the encoder-decoder model used in Smart Reply for Inbox.

One challenge we faced was that response generation in online conversations have very strict latency requirements. To address this, Pavel Sountsov and Sushant came up with an innovative two-stage model that works as follows. First, a recurrent neural network looks at the conversation context one word at a time and encodes it in the hidden state of a long short term memory (LSTM). Below, we show an example with a context ‘Where are you?’. The context has three tokens, each of which is embedded into a continuous space and input to the LSTM. The LSTM state now encodes the context as a continuous vector. This vector is used to generate the response as a discretized semantic class.
Each semantic class is associated with a set of possible messages that belong to it. We use a second recurrent network to generate a specific message from that set. This network also converts the context into a hidden LSTM state but this time the hidden state is used to generate the full message of the reply one token at a time. For example, now the LSTM after seeing the context “Where are you?” generates the tokens in the response: “I’m at work”.
A beam search is used to efficiently select the top-N highest scoring responses from among the very large set of possible messages that a LSTM can generate. A snippet of the search space explored by such a beam-search technique is shown below.
As with any large-scale product, there were several engineering challenges we had to solve in generating a set of high-quality responses efficiently. For example, in spite of the two staged architecture, our first few networks were very slow and required about half a second to generate a response. This was obviously a deal breaker when we are talking about real time communication apps! So we had to evolve our neural network architecture further to reduce the latency to less than 200ms. We moved from using a softmax layer to a hierarchical softmax layer which traverses a tree of words instead of traversing a list of words thus making it more efficient.

Another interesting challenge we had to solve when generating predictions is controlling for message length. Sometimes none of the most probable responses are appropriate - if the model predicts too short a message, it might not be useful to the user, and if we predict something too long, it might not fit on the phone screen. We solved this by biasing the beam search to follow paths that lead to higher utility responses instead of favoring just the responses that are most probable. That way, we can efficiently generate appropriate length response predictions that are useful to our users.

Personalized for you

The best part about these suggestions is that over time they are personalized to you so that your individual style is reflected in your conversations. For example, if you often reply to “How are you?” with “Fine.” instead of “I am good.”, it will learn your preference and your future suggestions will take that into account. This was accomplished by incorporating a user's "style" as one of the features in a Neural Network that is used to predict the next word in a response, resulting in suggestions that are customized for your personality and individual preferences. The user's style is captured in a sequence of numbers that we call the user embedding. These embeddings can be generated as part of the regular model training, but this approach requires waiting for many days for training to be complete and it cannot handle more than a handful of millions of users. To solve this issue, Alon Shafrir implemented a L-BFGS based technique to generate user embeddings quickly and at scale. Now, you'll be able to enjoy personalized suggestions after only a short time of using Allo.

More than just English

The neural network model described above is language agnostic so building separate prediction models for each language works quite well. To make sure that responses for each language benefit from our semantic understanding of other languages, Sujith Ravi came up with a graph-based machine learning technique that can connect possible responses across languages. Dana Movshovitz-Attias and Peter Young applied this technique to build a graph that connects responses to incoming messages and to other responses that have similar word embeddings and syntactic relationships. It also connects responses with similar meaning across languages based on the machine translation models developed by our Translate team.

With this graph, we use semi-supervised learning, as described in this paper, to learn the semantic meaning of responses and determine which are the most useful clusters of possible responses. As a result, we can allow the LSTM to score many possible variants of each possible response meaning, allowing the personalization routines to select the best response for the user in the context of the conversation. This also helps enforce diversity as we can now pick the final set of responses from different semantic clusters.

Here’s an example of how the graph might look for a set of messages related to greetings:
Beyond Smart Reply

I am also very excited about the Google assistant in Allo with which you can converse and get information about anything that Google Search knows about. It understands your sentences and helps you accomplish tasks directly from the conversation. For example, the Google assistant can help you discover a restaurant and reserve a table from within the Allo app when chatting with your friends. This has been made possible because of the cutting-edge research in natural language understanding that we have been doing at Google. More details to follow soon!

These smart features will be part of the Android and iOS apps for Allo that will be available later this summer. We can’t wait for you to try and enjoy it!

We wish to acknowledge the hard work of the following in building Smart Reply:

Ryan Cassidy, Dave Citron, Ori Gershony, Max Gubin, Pranav Khaitan, Harini Krishnamurthy, Patrick McGregor, Dana Movshovitz-Attias, Sergey Nazarov, Hung Pham, Sushant Prakash, Vivek Ramavajjala, Sujith Ravi, Sunita Sarawagi, Alon Shafrir, Pavel Sountsov, Peter Young, Shu Zhang