Embedded markup of Web pages has seen widespread adoption throughout the past years driven by standards such as RDFa and Microdata and initiatives such as schema.org, where recent studies show an adoption by 39% of all Web pages already in 2016. While this constitutes an important information source for tasks such as Web search, Web page classification or knowledge graph augmentation, individual markup nodes are usually sparsely described and often lack essential information. For instance, from 26 million nodes describing events within the Common Crawl in 2016, 59% of nodes provide less than six statements and only 257,000 nodes (0.96%) are typed with more specific event subtypes. Nevertheless, given the scale and diversity of Web markup data, nodes that provide missing information can be obtained from the Web in large quantities, in particular for categorical properties. Such data constitutes potential training data for inferring missing information to significantly augment sparsely described nodes. In this work, we introduce a supervised approach for inferring missing categorical properties in Web markup. Our experiments, conducted on properties of events and movies, show a performance of 79% and 83% F1 score correspondingly, significantly outperforming existing baselines.

Neural text generation models are often autoregressive language models or seq2seq models. These models generate text by sampling words sequentially, with each word conditioned on the previous word, and are state-of-the-art for several machine translation and summarization benchmarks. These benchmarks are often defined by validation perplexity even though this is not a direct measure of the quality of the generated text. Additionally, these models are typically trained via maxi- mum likelihood and teacher forcing. These methods are well-suited to optimizing perplexity but can result in poor sample quality since generating text requires conditioning on sequences of words that may have never been observed at training time. We propose to improve sample quality using Generative Adversarial Networks (GANs), which explicitly train the generator to produce high quality samples and have shown a lot of success in image generation. GANs were originally designed to output differentiable values, so discrete language generation is challenging for them. We claim that validation perplexity alone is not indicative of the quality of text generated by a model. We introduce an actor-critic conditional GAN that fills in missing text conditioned on the surrounding context. We show qualitatively and quantitatively, evidence that this produces more realistic conditional and unconditional text samples compared to a maximum likelihood trained model.

Thrilling to see more bots coming to the travel industry. This article does a great job outlining exactly what a good Bot Pilot structure should look like: Start with a simple use case/tackling your most FAQ. The purpose of the pilot isn't to get everything right, right off the bat, it's to learn and gather data to improve your bot's knowledge base and NLP. So structure your pilot with the intention of listening and identifying what people are asking it that you didn't anticipate. Be resourceful when it comes to tackling the discovery hurdle.

Welcome to the real "Minority Report". PRE-CRIME is a wake up call for all of us. A science fiction scenario, both fascinating and scary, has arrived in our everyday life. To predict a future crime scene and to prevent a murder seems like something from a sci-fi movie. It is, but it's also real – and happening right now.

Leonardo and David Ryan Polgar were running behind schedule for their show -- technical difficulties with the projector -- then came out for a cold open. "Tonight, we're talking about the future of work," said Polgar. "What are you worried about with the future of work, Joe?" "We're all gonna be replaced with robots," Leonardo shot back, looking half expectantly at the crowd of about 30 people, who looked back. Polgar moved on to the recent Boston Dynamics video, which bears a disturbing resemblance to the Black Mirror episode "Metalhead." A four-legged robot precisely opens a door for another robot, then holds it for the companion to enter. "They're already making me look bad, because there's, like chivalry in robots now," Leonardo said, to murmurs of chuckles.

Annihilation deals in bountiful hallucinogenic imagery, but the image from Alex Garland's sci-fi horror that may prove most remarkable to audiences is one that really ought to be mundane: a poster featuring the film's five female leads. It's an uncommon setup, and not just for a generously budgeted studio picture.

Annihilation deals in bountiful hallucinogenic imagery, but the image from Alex Garland's sci-fi horror that may prove most remarkable to audiences is one that really ought to be mundane: a poster featuring the film's five female leads. Female representation in Hollywood still lags far behind --women made up only 34 percent of speaking characters in top-grossing films last year, while the number of female leads has, in fact, recently fallen--but Natalie Portman, Jennifer Jason Leigh, Tessa Thompson, Gina Rodriguez, and Tuva Novotny have the reins of this $55 million Paramount project, while Oscar Isaac, the film's most significant male character, takes the supporting role of imperilled love interest to a take-action female hero. It's an uncommon setup, and not just for a generously budgeted studio picture. But it's less unusual when you narrow the focus to science fiction, where women have recently been taking the lead on-screen. Garland's sophomore effort as writer-director follows his own Ex Machina, plus such sizable productions as Arrival, Gravity, 10 Cloverfield Lane, The Cloverfield Paradox, Colossal, Okja, and The Shape of Water, in putting a woman or women at the forefront of a science-fiction narrative.

Artificial intelligence research has been around for more than half a century, but only in recent years have we seen developments in AI technology that might bring sci-fi film plotlines to life. Merriam-Webster defines AI as "a branch of computer science dealing with the simulation of intelligent behavior in computers" and "the capability of a machine to imitate intelligent human behavior." Today, the more benign uses of AI include software that scans thousands of job applicant resumes in minutes – saving recruiters countless hours of manual screening – and AI algorithms that review patient X-rays and CT scans to highlight potential findings for radiologists. But even the beneficial aspects of AI – increased efficiency and accuracy – portend an increasingly automated job market in which some workers will continue to see their positions replaced by machines. Some of the more nefarious applications of AI have been the stuff of Hollywood movies for years – like the prospect of widespread death and destruction caused by unmanned autonomous weapons designed to make lethal decisions.

Hashtag creator Chris Messina today launched Molly, a service that allows people to ask questions about you and glean information from your various social media profiles. Molly skims your posts on platforms like Instagram, Twitter, and Medium to learn about you and formulate natural language questions. When someone asks something Molly can't answer, that question is sent to the Molly app for you to answer yourself. In addition to following your social media activity, the Molly app asks you to answer questions about yourself, like "Do you own an Amazon Echo?" or "Do you have a sweet tooth or a savory tooth?" The more you swipe through the questions, the more Molly learns about you, and the more you learn about how your friends have answered similar questions.

Last fall, Amazon CEO Jeff Bezos charged his company's studio division to produce bigger shows with a "global appeal," in an attempt to compete with other streaming video companies in the race to produce high-end original content. Since then, Amazon has moved quickly to begin developing an impressive slate of genre television shows that include adaptations of novels such as Ringworld, Snow Crash, and J.R.R. Tolkien's Lord of the Rings. Now, The Hollywood Reporter reports that Amazon has acquired the rights to another huge book: Consider Phlebas, the first installment of Iain M. Banks's space opera Culture series. THR reports that Amazon will partner with British screenwriter Dennis Kelly (who created the critically acclaimed Channel 4 series Utopia), and that should the scripts for the series work out, it will order it directly to a series. By acquiring big-name franchises such as Lord of the Rings, while facing intense competition from companies like Netflix, Apple, and Disney, it's clear that Amazon is working to aggressively pursue major projects that play with complicated stories.