With more than 2 billion users, Facebook Inc. has to deal with dozens of different languages on its social network, which poses a bit of a barrier to the company's new mission to "bring the world closer together." Facebook hopes that artificial intelligence will be the answer to this problem, and today the company announced that its translations now rely entirely on cutting-edge neural machine learning. In a blog post published today, Facebook researchers Juan Miguel Pino, Alexander Sidorov and Necip Fazil Ayan explained just how hard it is to deal with so many languages. "Creating seamless, highly accurate translation experiences for the 2 billion people who use Facebook is difficult," they said. "We need to account for context, slang, typos, abbreviations, and intent simultaneously."

Facebook has been working on changing how it translates text in posts and comments and today it announced that its transition is complete. It means that translations should be quite a bit more accurate going forward. Previously, Facebook was using phrase-based machine translation models, which break down sentences into words or phrases, limiting how they can go about translating a full sentence. These sorts of models' shortcomings were particularly evident when translating between languages with really different sentence structures. Now, however, the site is using neural networks to power its translations, which can take into account full sentences as well as their context, generating much more accurate translations.

Well, Microsoft (via Skype Translate), Google (Translate), and others have done this and beyond. Ray's predictions are a byproduct of his (and my) understanding of the power of Moore's Law, more specifically Ray's "Law of Accelerating Returns" and of exponential technologies. Before we know it, they are DISRUPTIVE--just look at the massive companies that have been disrupted by technological advances in AI, virtual reality, robotics, internet technology, mobile phones, OCR, translation software, and voice control technology. Now, these technologies power multibillion-dollar companies and affect billions of lives.

In its latest paper, the Facebook AI Research (FAIR) team dropped some impressive results for its implementation of a modified convolutional neural network for machine translation. Facebook says it has achieved a small bump in accuracy at nine times the speed of traditional recurrent network models. And to complement its research, the company is releasing its pre-trained models on GitHub, along with all the tools needed to replicate the results on your own. When most of us think of machine translation, we think of Google Translate (sorry Facebook and my 8th grade Spanish teacher). But while that is certainly the most well-known implementation, Facebook relies on the technology extensively for translating posts on News Feed, among other uses.

We introduce and describe the results of a novel shared task on bandit learning for machine translation. The task was organized jointly by Amazon and Heidelberg University for the first time at the Second Conference on Machine Translation (WMT 2017). The goal of the task is to encourage research on learning machine translation from weak user feedback instead of human references or post-edits. On each of a sequence of rounds, a machine translation system is required to propose a translation for an input, and receives a real-valued estimate of the quality of the proposed translation for learning. This paper describes the shared task's learning and evaluation setup, using services hosted on Amazon Web Services (AWS), the data and evaluation metrics, and the results of various machine translation architectures and learning protocols.

"If more visitors from overseas come to Japan, it's possible that the number of (non-Japanese) inmates will increase," Dai Tanaka, an official in the ministry's prison services division, said while providing one reason for introducing the new video phone service. Since the number of approved translators are limited, prisons have sometimes faced challenges when responding to visitation requests by inmates' families, Tanaka said. Language experts at Fuchu Prison have been testing the new video phone translation system since last August. According to Tanaka, there are currently 76 major detention facilities in the nation, including prisons, facilities for juveniles and detention centers.

We propose an approach to build a neural machine translation system with no supervised resources (i.e., no parallel corpora) using multimodal embedded representation over texts and images. Based on the assumption that text documents are often likely to be described with other multimedia information (e.g., images) somewhat related to the content, we try to indirectly estimate the relevance between two languages. Using multimedia as the "pivot", we project all modalities into one common hidden space where samples belonging to similar semantic concepts should come close to each other, whatever the observed space of each sample is. This modality-agnostic representation is the key to bridging the gap between different modalities. Putting a decoder on top of it, our network can flexibly draw the outputs from any input modality. Notably, in the testing phase, we need only source language texts as the input for translation. In experiments, we tested our method on two benchmarks to show that it can achieve reasonable translation performance. We compared and investigated several possible implementations and found that an end-to-end model that simultaneously optimized both rank loss in multimodal encoders and cross-entropy loss in decoders performed the best.

How artificial intelligence got shaped during the history? What is the role of big data in A.I? Big data contains thе quantity аnd diversity оf high frеquеnсу digital dаtа. Big data consists of Internet, Meta data: tags, translations and mechanical Turk. Big Data on its own is not a useful thing. It's bunch of information unless you apply a methodology to make use of it.

We propose a novel decoding approach for neural machine translation (NMT) based on continuous optimisation. We convert decoding - basically a discrete optimization problem - into a continuous optimization problem. The resulting constrained continuous optimisation problem is then tackled using gradient-based methods. Our powerful decoding framework enables decoding intractable models such as the intersection of left-to-right and right-to-left (bidirectional) as well as source-to-target and target-to-source (bilingual) NMT models. Our empirical results show that our decoding framework is effective, and leads to substantial improvements in translations generated from the intersected models where the typical greedy or beam search is not feasible. We also compare our framework against reranking, and analyse its advantages and disadvantages.