Week Ending 07.21.19
RESEARCH WATCH: 07.21.19
Over the past week, 823 new papers were published in "Computer Science".
The paper discussed most in the news over the past week was "Tracking sex: The implications of widespread sexual data leakage and tracking on porn websites" by Elena Maris et al (Jul 2019), which was referenced 188 times, including in the article Facebook and Google know what porn you're watching, even when you're in incognito in Business Insider. The paper author, Elena Maris (Postdoctoral researcher at Microsoft), was quoted saying "The fact that the mechanism for adult site tracking is so similar to, say, online retail should be a huge red flag". The paper got social media traction with 90 shares. The researchers explore tracking and privacy risks on pornography websites. A user, @citadelo, tweeted "This research focuses on porn-sites users' tracking: "analysis of 22,484 pornography websites indicated that 93% leak user data to a third party." Unfortunately even incognito window does not solve the entire problem. Another reason to use".
Leading researcher Ruslan Salakhutdinov (Carnegie Mellon University) came out with "Learning Neural Networks with Adaptive Regularization".
The paper shared the most on social media this week is "What does it mean to understand a neural network?" by Timothy P. Lillicrap et al (Jul 2019) with 526 shares. @Kleinspaces (David J Klein) tweeted "For decades we’ve been making inspiring but weak comparisons between the brain & artificial systems learned from simple principles. Time to formalize the study of learning rules/objectives/constraints in neuroscience. This article argues this is what neuroscience should become".
Over the past week, 49 new papers were published in "Computer Science - Artificial Intelligence".
The paper discussed most in the news over the past week was by a team at Massachusetts Institute of Technology: "Tackling Climate Change with Machine Learning" by David Rolnick et al (Jun 2019), which was referenced 22 times, including in the article How artificial intelligence can tackle climate change in National Geographic. The paper author, David Rolnick (Massachusetts Institute of Technology), was quoted saying "Climate change does not present one problem, it presents multiple problems. AI is only one of the tools that can have an impact in the fight to mitigate the effects of climate change". The paper also got the most social media traction with 1459 shares. The researchers describe how machine learning can be a powerful tool in reducing greenhouse gas emissions and helping society adapt to a changing climate. On Twitter, @ikdeepl said "Yikes “In 2006, at least two Scottish seafood firms flew hundreds of metric tons of shrimp from Scotland to China and Thailand for peeling, then back to Scotland for sale – because they could save on labor costs”".
Leading researcher Sergey Levine (University of California, Berkeley) came out with "Dynamical Distance Learning for Unsupervised and Semi-Supervised Skill Discovery".
Over the past week, 163 new papers were published in "Computer Science - Computer Vision and Pattern Recognition".
The paper discussed most in the news over the past week was by a team at UC Berkeley: "Natural Adversarial Examples" by Dan Hendrycks et al (Jul 2019), which was referenced 12 times, including in the article If you can identify what’s in these images, you’re smarter than AI in The Verge. The paper got social media traction with 509 shares. A Twitter user, @DanHendrycks, posted "Natural Adversarial Examples are real-world and unmodified examples which cause classifiers to be consistently confused. The new dataset has 7,500 images, which we personally labeled over several months. Paper: Dataset and code".
Leading researcher Sergey Levine (University of California, Berkeley) published "Dynamical Distance Learning for Unsupervised and Semi-Supervised Skill Discovery".
This week was active for "Computer Science - Computers and Society", with 31 new papers.
The paper discussed most in the news over the past week was "Tracking sex: The implications of widespread sexual data leakage and tracking on porn websites" by Elena Maris et al (Jul 2019)
The paper shared the most on social media this week is "Data-driven strategies for optimal bicycle network growth" by Luis Natera et al (Jul 2019) with 100 shares. @alexvespi (Alessandro Vespignani) tweeted "Data-driven strategies for optimal bicycle network growth “algorithms add the most critical missing links in the bicycle layer: The first algorithm connects the two largest connected components, the second connects the largest with the closest component”".
This week was active for "Computer Science - Human-Computer Interaction", with 33 new papers.
The paper discussed most in the news over the past week was by a team at University of Rochester: "Discourse Behavior of Older Adults Interacting With a Dialogue Agent Competent in Multiple Topics" by S. Zahra Razavi et al (Jul 2019), which was referenced 2 times, including in the article Study: Seniors talk with AI chatbots more when the conversation is deeper | VentureBeat in Venturebeat. The paper got social media traction with 10 shares.
The paper shared the most on social media this week is by a team at Google: "The Bach Doodle: Approachable music composition with machine learning at scale" by Cheng-Zhi Anna Huang et al (Jul 2019) with 186 shares. @joe_antognini (Joseph O'Brien Antognini) tweeted "Google AI resident Anna Huang has a really cool paper detailing the ML behind the Bach Google doodle that appeared back in March! They're also releasing the dataset of 20 million melodies that users composed".
This week was very active for "Computer Science - Learning", with 283 new papers.
The paper discussed most in the news over the past week was by a team at Massachusetts Institute of Technology: "Tackling Climate Change with Machine Learning" by David Rolnick et al (Jun 2019)
Leading researcher Ruslan Salakhutdinov (Carnegie Mellon University)
The paper shared the most on social media this week is "What does it mean to understand a neural network?" by Timothy P. Lillicrap et al (Jul 2019)
Over the past week, nine new papers were published in "Computer Science - Multiagent Systems".
The paper discussed most in the news over the past week was by a team at University of Tartu: "Perspective Taking in Deep Reinforcement Learning Agents" by Aqeel Labashet al (Jul 2019), which was referenced 2 times, including in the article Last Week in AI in Hacker Noon. The paper got social media traction with 23 shares. The researchers present their progress toward building artificial agents with such abilities. A Twitter user, @dawn_alderson, posted "yes, a meta-learning strategy to close the gap between level 0 and level 2 perspective taking=RL has much promise but, I rather suspect hybridity needs to evolve in terms of the LSTM stack. hh", while @jaaanaru observed "Our first step toward creating AI agents with theory of mind: deep reinforcement learning agents can learn to take the perspective of another agent. This work was inspired by experiments done with chimpanzees, so we have monkeys and bananas on figures!".
Over the past week, 28 new papers were published in "Computer Science - Neural and Evolutionary Computing".
The paper discussed most in the news over the past week was by a team at Massachusetts Institute of Technology: "A Power Efficient Artificial Neuron Using Superconducting Nanowires" by Emily Toomey et al (Jun 2019), which was referenced 2 times, including in the article Superconducting neurons could match the power efficiency of the brain in Technology Review. The paper got social media traction with 8 shares. A Twitter user, @mehrdad_hp_, posted "Make a worm first! o_O".
The paper shared the most on social media this week is "Are We Really Making Much Progress? A Worrying Analysis of Recent Neural Recommendation Approaches" by Maurizio Ferrari Dacrema et al (Jul 2019) with 178 shares. The authors report the results of a systematic analysis of algorithmic proposals for top - n recommendation tasks. @ekshakhs (ni sinha) tweeted ""Consider 18 algorithms that were presented at top-level research conferences in the last years. Only 7 of them could be reproduced with reasonable effort... 6 of them can often be outperformed with comparably simple heuristic methods". Systemic problem with bleeding research!".
This week was active for "Computer Science - Robotics", with 46 new papers.
The paper discussed most in the news over the past week was by a team at University of North Carolina: "Identifying Emotions from Walking using Affective and Deep Features" by Tanmay Randhavane et al (Jun 2019), which was referenced 7 times, including in the article Indian Researcher & His Team Build AI That Can Tell How We Feel Just By Seeing Us Walk in Indiatimes. The paper author, Aniket Bera (University of North Carolina), was quoted saying "Though we do not make any claims about the actual emotions a person is experiencing, our approach can provide an estimate of the perceived emotion of that walking style". The paper got social media traction with 191 shares. On Twitter, @joemmac said "Gait is also sufficiently individualized that it can be used to ID targets with far less resolution and perfect angles than facial detection. And yes I used "targets" on purpose... welcome to the dystopia!".
Leading researcher Sergey Levine (University of California, Berkeley)