ACM TechNews
Welcome to the March 27, 2017 edition of ACM TechNews, providing timely information for IT professionals three times a week.

ACM TechNews mobile apps are available for Android phones and tablets (click here) and for iPhones (click here) and iPads (click here).

traffic jam New AI Algorithm Beats Even the World's Worst Traffic
Motherboard
Michael Byrne
March 27, 2017


Researchers at Nanyang Technological University (NTU) in Singapore have developed an intelligent routing algorithm designed to minimize the occurrence of spontaneous traffic jams across a roadway network. The fast, computationally distributed algorithm begins by assuming breakdowns will occur with enough traffic density. "Our goal is to direct the traffic flow so that the overall traffic breakdown probability is minimized," write NTU's Hongliang Guo and colleagues. They say in another context, "our objective is to maximize the probability that none of the network links encounters a traffic breakdown." The problem then becomes one handled by machine learning, which yields optimal routes through the network. Guo's team has devised mathematical optimizations that make real-time calculation feasible, and they also demonstrated their algorithm in simulated situations. The researchers determined only 10 percent of cars in a network should be driving for those optimizations to have a positive effect on the network.

Full Article
It's Time for Canada to Invest in Developing Artificial Intelligence
Toronto Globe & Mail (Canada)
Alan Bernstein; Pierre Boivin; David McKay
March 26, 2017


Canada's federal government this past week unveiled an artificial intelligence (AI) initiative that will establish the foundation for a Pan-Canadian Artificial Intelligence Strategy, according to Canadian Institute for Advanced Research (CIFAR) CEO Alan Bernstein, Claridge CEO Pierre Boivin, and Royal Bank of Canada CEO David McKay. Through CIFAR, the initiative will help three AI institutes enlist and retain scientific talent and train hundreds of graduate students while also funding research into the social, legal, and ethical ramifications of deep AI. The goal is to cultivate a Canadian brand of AI technology that serves human needs, problems, and ambitions. However, Bernstein, Boivin, and McKay agree Canadian AI research is losing ground to bigger global competitors. They say what is needed is prioritization of expanding the country's talent pipeline, fostering conditions under which AI entrepreneurs can thrive, and helping entrenched businesses leverage AI.

Full Article

lemonade Virtual Lemonade Sends Color and Taste to a Glass of Water
New Scientist
Timothy Revell
March 24, 2017


Researchers at the National University of Singapore say they have developed a system of sensors and electrodes that can digitally transmit the basic color and sourness of a glass of lemonade to a tumbler of water, enabling people to share sensory experiences over the Internet. Nimesha Ranasinghe and colleagues used an RBG color sensor and a pH sensor to capture the color and acidity of a freshly poured glass of lemonade. They then transmitted this data to a special tumbler equipped with an electrode that mimicked the sourness of the lemonade by stimulating the drinker's taste buds with a pulse of electricity. Light-emitting diodes were used to replicate the color of the drink. Ranasinghe says these types of virtual flavorings could help people enjoy digital versions of sugary drinks without consuming calories or damaging their teeth.

Full Article
Parallel Computation Provides Deeper Insight Into Brain Function
Okinawa Institute of Science and Technology
Greta Keenan
March 24, 2017


Researchers at the Okinawa Institute of Science and Technology (OIST) in Japan have developed Parallel STEPS, new software that models neurons with the goal of discovering new insights into neuronal function. The Parallel STEPS software is based on two related studies, one of which focused on ensuring that the accuracy of the software is comparable with conventional methods, while the other investigated the software's performance and potential applications. The findings show Parallel STEPS implementation achieves significant improvements in performance, and good scalability when compared to similar models. Although STEPS is only realistically capable of modeling parts of neurons, the researchers hope to develop a full-scale model of a whole neuron as well as the interactions between neurons in a network. The OIST team says they will work with colleagues from Swiss Federal Institute of Technology in Lausanne, Switzerland and will utilize the IBM Blue Gene/Q supercomputer to achieve these goals.

Full Article

Women in Computing tutoring event What It Means to {Codelikeagirl}
University of Rochester NewsCenter
Bob Marcotte
May 20, 2017


The University of Rochester has significantly increased the number of female students graduating with computer science degrees, and this year it will be double the national average. The university accomplished this partly with a cluster system enabling students to sample computer science courses and then opt to continue with a minor or double major. Another contributing factor has been the participation of Rochester's computer science department in the Building Recruiting And Inclusion for Diversity (BRAID) initiative. In addition to ramping up the department's recruitment of females and minorities and creating more accessible courses for clusters and majors, BRAID underwrites student attendance at the Grace Hopper Celebration of Women in Computing, where networking and mentoring opportunities are abundant. The event encourages a communal feeling among attendees, which has been a catalyst for Women in Computing, a subcommittee of the Computer Science Undergraduate Council.

Full Article
Machine Learning Opens Up New Ways to Help Disabled People
Technology Review
Tom Simonite
March 23, 2017


IBM and other companies are developing and launching machine-learning technologies and applications to enhance the experience of disabled persons. For example, IBM researchers are using language-processing software developed under the company's Watson project to support Content Clarifier, a tool designed to help cognitively- or intellectually-challenged people. It can substitute figures of speech such as "raining cats and dogs" with plainer terms, and shorten or break up long sentences with multiple clauses and indirect language. IBM researcher Will Scott says his company is in discussions with a organization that helps autistic high school students migrate to college life about testing Content Clarifier to help people comprehend educational and administrative documents. Tackling the reliability challenge of such machine-learning tools is essential if they are to be successful, according to Artist with Autism participant Austin Lubetkin, who has prototyped a navigation app using image-recognition technology to offer directions in the form of landmarks.

Full Article
These AI Bots Created Their Own Language to Talk to Each Other
ReCode
April Glaser
March 23, 2017


OpenAI last week disclosed new research detailing the training of artificial intelligence (AI) bots to generate their own shared language based on reinforcement learning as they navigate a set environment. OpenAI scientists tasked the AIs, represented as differently colored circles, with accomplishing specific goals in collaboration by communicating in their own language. The bots produced numerical terms corresponding directly with environmental objects and other bots and actions, which the researchers tagged with English words. The AIs were taught how to communicate via trial and error, so they could recall successful and unsuccessful strategies the next time they were told to complete a task. "We think that if we slowly increase the complexity of their environment, and the range of actions the agents themselves are allowed to take, it's possible they'll create an expressive language which contains concepts beyond the basic verbs and nouns that evolved here," the researchers note.

Full Article

A game on a mobile phone An Algorithm That Knows When You'll Get Bored With Your Favorite Mobile Game
Platforma Sinc
March 23, 2017


An international team of researchers has developed a new algorithm, called survival ensemble, that predicts when a user will leave a mobile game. The mathematical model uses the "ensemble" method, which is based on many learning algorithms instead of a single one, thereby improving the prediction accuracy by examining many more correlations and alternative models. The researchers also used a survival analysis algorithm with each submodel, which are traditionally used in medical research. The team combined the power of survival algorithms and ensemble models to analyze video game interest. This technique enabled the researchers to achieve a high level of prediction accuracy, as the algorithm automatically adapts to the data of the game being analyzed. The researchers say when applied to video games, the model can predict what day and at what stage of the game a user will stop playing, and why they will do so.

Full Article
More Open to Coercion
A*STAR Research
March 22, 2017


A*STAR researchers in Singapore have devised a material that adds rhodium to enhance the properties of cobalt iridium, which could lower the power consumption of magnetic memories. Via direct current magnetron sputtering, the team injected cobalt, iridium, and rhodium simultaneously from separate solid sources in a vacuum chamber and deposited on a silicon substrate. Changing the energy supplied to each of the sources enabled the researchers to control the composition of the final material, boosting the amount of rhodium while subtracting iridium. Measurements of the magnetic properties of CoIr-Rh films showed the introduction of rhodium reduced the coercivity and the damping constant by more than 50 percent of that of unmodified cobalt iridium. "Such negative magnetocrystalline anisotropy materials enable higher frequency operation at lower driving current and the creation of a higher in-plane alternating-current magnetic field for effective assisted switching, and higher stability against stray fields and temperature fluctuations," says A*STAR's Tiejun Zhou.

Full Article

Researchers hold component of the Neuromorphic Cyber Microscope New Brain-Inspired Cybersecurity System Detects 'Bad Apples' 100 Times Faster
Sandia Labs News
Mollie Rappe
March 21, 2017


Researchers at Sandia National Laboratories and Lewis Rhodes Labs have designed a brain-inspired cybersecurity system that can spot attack indicators by analyzing the complex patterns underlying specific malefactors while consuming less electricity than a 60-watt lightbulb. The Neuromorphic Cyber Microscope is modeled after the brain's ability to continuously scan for threats, comparing streaming data to suspicious patterns in a time-dependent fashion. Sandia's Roger Suppona says the new system is more than 100 times faster and 1,000 times more energy-efficient than racks of conventional cybersecurity systems. "This completely changes the way that we look for suspicious activity without running the risk of overwhelming our analysts with too much information," he notes. The collaborators also are exploring other applications for general neuromorphic architecture, including a type of machine learning for audio and image processing and sorting numbers efficiently. The current focus of this exploration is basic algorithms, according to Sandia's John Naegle.

Full Article

A trio of mobile apps Apps Created for Everyone, by Anyone
MIT News
Rob Matheson
March 21, 2017


Modo Labs is building on its Kurogo platform, which simplifies application-building by providing fast and easy access to back-end data and a simple mobile app-assembly interface, for wide adoption by universities, firms, and other clients. Modo founder Andrew Yu says the simplicity of the app-builder is rooted in optimization innovations such as synchronization across multiple mobile devices, operating systems, and screen sizes, and mobile-optimized middleware. Modo's genesis began during Yu's tenure at the Massachusetts Institute of Technology (MIT), in developing the open source MIT Mobile Framework program. Yu says friendly competition between Modo's customers is spurring his startup to constantly reinvent its app-building tool. The Modo platform is currently in use by 200 universities and colleges, along with Fortune 500 companies and hospitals across the globe. Yu's near-future plans include the platform's adoption by 1,000 higher education institutions and other organizations.

Full Article

A serene scene from ArcheAge video game People Remain Calm as the World Ends, Video Game Study Suggests
University at Buffalo News
Cory Nealon
March 20, 2017


University at Buffalo researchers found despite some violent acts, most people will be helpful to others as the world ends. This conclusion is based upon the virtual actions of more than 80,000 players of the video game "ArcheAge." The researchers argue despite the obvious limitations of the study--namely that it is based upon a video game, not real life--it offers a realistic view into the behavior of people in an end-times scenario. As part of the study, the researchers analyzed 275 million records of player behavior that were recorded during a trial of "ArcheAge" before the medieval fantasy game was released to the public in January 2013. The researchers determined although a small percentage of the overall population did participate in anti-social behavior, most players exhibited pro-social behavior such as strengthening existing social relationships and forming new ones.

Full Article
ACM Queue Mobile Apps
 
ACM Conferences
 

Association for Computing Machinery

2 Penn Plaza, Suite 701
New York, NY 10121-0701
1-800-342-6626
(U.S./Canada)



ACM Media Sales

If you are interested in advertising in ACM TechNews or other ACM publications, please contact ACM Media Sales or (212) 626-0686, or visit ACM Media for more information.