In neural abstractive summarization field, conventional sequence-to-sequence based models often suffer from summarizing the wrong aspect of the document with respect to the main aspect. To tackle this problem, we propose the task of reader-aware abstractive summary generation, which utilizes the reader comments to help the model produce better summary about the main aspect. Unlike traditional abstractive summarization task, reader-aware summarization confronts two main challenges: (1) Comments are informal and noisy; (2) jointly modeling the news document and the reader comments is challenging. To tackle the above challenges, we design an adversarial learning model named reader-aware summary generator (RASG), which consists of four components: (1) a sequence-to-sequence based summary generator; (2) a reader attention module capturing the reader focused aspects; (3) a supervisor modeling the semantic gap between the generated summary and reader focused aspects; (4) a goal tracker producing the goal for each generation step. The supervisor and the goal tacker are used to guide the training of our framework in an adversarial manner. Extensive experiments are conducted on our large-scale real-world text summarization dataset, and the results show that RASG achieves the state-of-the-art performance in terms of both automatic metrics and human evaluations. The experimental results also demonstrate the effectiveness of each module in our framework. We release our large-scale dataset for further research.

Deep learning (DL) defines a data-driven programming paradigm that automatically composes the system decision logic from the training data. In company withthe data explosion and hardware acceleration during the past decade, DL achieves tremendous success in many cutting-edge applications. However, even the state-of-the-art DL systems still suffer from quality and reliability issues. It was only until recently that some preliminary progress was made in testing feed-forward DL systems.In contrast to feed-forward DL systems, recurrent neural networks(RNN) follow a very different architectural design, implementing temporal behaviours and "memory" with loops and internal states. Such stateful nature of RNN contributes to its success in handling sequential inputs such as audio, natural languages and video processing, but also poses new challenges for quality assurance. In this paper, we initiate the very first step towards testing RNN-based stateful DL systems. We model RNN as an abstract state transition system, based on which we define a set of test coverage criteria specialized for stateful DL systems. Moreover, we propose an automated testing framework, DeepCruiser, which systematically generates tests in large scale to uncover defects of stateful DL systems with coverage guidance. Our in-depth evaluation on a state-of-the-art speech-to-text DL system demonstrates the effectiveness of our technique in improving quality and reliability of stateful DL systems.

A slew of rumours surrounding the Galaxy S10 mean there are few secrets left to reveal about Samsung's highly-anticipated flagship smartphone. One unknown that still remains, however, is the release date and price of the Galaxy S9 successor and its variants. Industry speculation has so far suggested the Mobile World Congress conference in Barcelona on 25 February could be a likely launch date, but a fresh leak claims it could come even sooner. The Galaxy S10 will instead be unveiled at a Samsung event ahead of MWC, with pre-orders beginning from this yet-to-be revealed date. The smartphone will then be released on 8 March 2019.

A robot hailed as the most advanced created in Russia was actually just a man in a suit. The android – known as Boris – took to the stage at a Russian technology conference to delight the world with its entirely lifelike moving and dancing. It was so impressive that it appeared on Russian state television, celebrated as the most modern robot, and produced by a team of students. Video of the event went around the world, showing him taking part in banter with people on stage and being led through a series of dances. Its success was used to encourage children to explore robotics, and as proof of a technological breakthrough.

The Foreign Ministry plans to hold Japanese-language tests in at least seven countries for a new type of work visa, in line with the introduction of the new visa category in April, sources have said. The ministry also aims to conclude bilateral agreements with those countries by March to share information related to criminal investigations, in an effort to shut out rogue labor brokers and protect workers' human rights, according to the sources. The Japanese-language tests will be administered in Vietnam, China, the Philippines, Indonesia, Thailand, Myanmar and Cambodia for applicants seeking the easier of the two new visa categories, the sources said. The government is negotiating with one more country on the matter, and plans to expand the number of countries where the tests are conducted in stages, according to the sources. The moves indicate that workers from those countries will be among the first group to be accepted in Japan under the new system, for the time being, aside from interns transferring from the existing foreign trainee program.

A robot on show at a Russian state-sponsored event has turned out to be a man dressed in a costume. Robot Boris featured on Russian TV and was apparently able to walk, talk and dance. But soon after its appearance journalists began to question the bot's authenticity. In a picture published afterwards on social media, the neck of a person was clearly visible. The robot is in fact a 250,000 rouble (£2,975) costume called Alyosha the Robot, made by a company called Show Robots. While the organisers of the Proyektoria technology forum - which was aimed at youngsters - had not claimed the robot was real, the TV coverage on Russia-24 suggested it was.

If you have been tempted to solve the numerous IQ tests that have plagued the social platforms, then you are not alone. Of late, emotional intelligence has been quantified for centuries through psychological tests and examinations that test your intelligence, personality, vocational interests, attitude, achievement, aptitude, and observational powers. It was never so easy to capture emotional intelligence as it is now. Human behaviour can be predicted by understanding and calculating one's emotions. AI startups have developed the technology where facial emotions and micro-expressions that lasted for seconds are studied and examined to establish patterns, then evaluated to find out what provokes these emotions, to conclusively determine underlying behaviour traits.

New rules mean drones are set to take off in India at last. But the legislation also requires pilots of all but tiny craft to get approval for every flight, even a zip around the local park. The regulations came into force earlier this month, overturning a 2014 ban on drone use by anyone other than government agencies. Now, commercial and recreational drones can fly during daytime, as long as they stay below 120 metres and within sight of operators.

Google's chief executive, Sundar Pichai, testified before the House judiciary committee on Tuesday morning, three months after his company thumbed its nose at Congress by failing to appear alongside Facebook and Twitter at a Senate hearing on election interference. In a hearing heavy on partisan theatrics, Pichai notably refused to rule out launching a censored search engine in China, a controversial plan that has garnered significant criticism from human rights organizations as well as rank-and-file Google employees. "Right now there are no plans to launch search in China," Pichai said numerous times, repeating a talking point that the company has relied on since news of the project leaked in August. Pichai characterized the Chinese search product as an "internal effort" and said the company would be "transparent" and consult with policy makers before launching in China. Pressed to rule out launching a tool that would enable censorship and surveillance in China, however, Pichai appeared to offer the company's probable justification for reentering a market that it left in 2010: "We think it's in our duty to explore possibilities to give users access to information."

Google has revealed what people in the UK have been searching for in 2018, with the publication of its annual'Year in Search' results. Meghan Markle, who topped the list last year, was knocked off her perch by the World Cup but still managed to be the second-most searched term in 2018. Her marriage to Prince Harry, together with the wedding of Princess Eugenie and Jack Brooksbank, helped'Royal Wedding' reach third on the overall list. Several other celebrities like Stormy Daniels and Ant McPartlin also appeared on the lists, which covered a wide range of categories including top trending'How to...?' and'What is...?' queries. Privacy scandals helped'How to delete Facebook?' become one of the most-asked questions using Google, while the dramatic rise and fall in cryptocurrency prices helped drive searches for'What is bitcoin?'.