Association for Computing Machinery
Welcome to the August 8, 2016 edition of ACM TechNews, providing timely information for IT professionals three times a week.

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


Stanford-Led Experiments Point Toward Memory Chips 1,000 Times Faster Than Today's
Stanford News (08/08/16) Tom Abate

Experiments led by Stanford University are exploring a new class of semiconductor materials that could form the basis of phase-change memory capable of permanent and faster data storage. "This work is fundamental but promising," says Stanford professor Aaron Lindenberg at the SLAC National Accelerator Laboratory. "A thousandfold increase in speed coupled with lower energy use suggests a path toward future memory technologies that could far outperform anything previously demonstrated." Phase-change materials can exist in two different atomic structures, and they preserve whatever electronic state conforms to their structure. Once their atoms flip or flop to form a 1 or a 0, the material stores that data until another burst of energy induces change. The researchers measured the intermediate phase when the material's amorphous structure, which inhibits electron flows, starts switching to a crystalline arrangement, determining it occurred less than a picosecond after the electrical jolt. The research suggests phase-change memory could greatly overtake silicon random-access memory's speed for tasks requiring memory-processor interoperation for computation. The researchers also say the experimental results imply such materials could execute superfast memory operations and permanent storage, depending on the duration of the thermal excitation's retention within the memory.

Cyber Protections Contemplated for U.S. Election Systems
Federal Computer Week (08/05/16) Mark Rockwell

Following repeated hacks of Democratic National Committee systems by attackers who could be associated with the Russian government, the Obama administration is considering boosting cyber protections for U.S. election systems by classifying them as critical infrastructure, which would put them under the protection of the U.S. Department of Homeland Security (DHS). "We have to carefully consider whether our election system is critical infrastructure, like the financial system or the power grid," says DHS secretary Jeh Johnson. Presidential assistant Lisa Monaco says the reaction to those who hack election systems in the U.S. might resemble what happened in response to the cyberattack on Sony Pictures Entertainment, which crossed a threshold into being destructive and coercive. She notes the U.S. government attributed the Sony attack to North Korea and hit the country with sanctions. In addition, the government also prosecuted Chinese military personnel who hacked into U.S. companies' systems to steal data, and recently indicted Iranian hackers for a series of cyberattacks. Monaco says a deliberate intrusion to coerce or influence the U.S. political process is a "serious, serious issue," which could require a new type of response.

Machine-Learning Algorithm Combs the Darknet for Zero Day Exploits, and Finds Them
Technology Review (08/05/16)

Researcher Eric Nunes and colleagues at Arizona State University have developed a machine-learning algorithm designed to identify zero-day exploits before they can be turned into malware. They say the algorithm scours hacking forums and marketplaces in the dark Web and deep net for hints of such vulnerabilities' existence. "Currently, this system collects on average 305 high-quality cyberthreat warnings each week," the researchers note. They designed a crawler to collect information from HTML pages hosted on the deep net and the darknet, with humans directing it to the best starting pages and then filtering out specific information about hacking activities. The researchers then employed the machine-learning algorithm to spot relevant products and topics discussed on these sites by labeling 25 percent of the data manually, citing what was relevant and what was not. The next step was to train the algorithm using this dataset and test it on the rest. "With the use of machine learning models, we are able to recall 92 percent of products in marketplaces and 80 percent of discussions on forums relating to malicious hacking with high precision," the researchers say. They report the method identified 16 zero-day exploits from the marketplace data over four weeks, and also mapped the social networks to learn how hackers use these forums and marketplaces.

DARPA: Autonomous Bug-Hunting Bots Will Lead to Improved Cybersecurity
U.S. Department of Defense (08/07/16) Cheryl Pellerin

Seven finalist teams last week at the U.S. Defense Advanced Research Projects Agency's (DARPA) Cyber Grand Challenge (CGC) demonstrated machines can autonomously find and patch software bugs. DARPA director Arati Prabhakar cites more than 200 active DARPA programs and the agency's mission "to change what's possible so that we can take huge strides forward in our national security capabilities." She describes DARPA's next grand challenge as a contest that tasks teams to build radio networks with embedded artificial intelligence (AI), which will enable each of the networks to dynamically scan and form theories and anticipate what is happening in radio spectrum. Prabhakar says through competition and collaboration, AI networks can advance the amount of capacity available from a fixed amount of spectrum. "As we start developing this technology and as we start implementing it...across the entire information ecosystem...we can start imagining a future where we actually have some foundation of cybersecurity," she notes. DARPA program manager Mike Walker says the prototypes tested via the CGC function on a simple research operating system and consume a huge amount of power in thinking about the security problems of small example services. He notes the bugs they detected during the CGC lack the complexity of real-world bugs, and much engineering must be conducted before such prototypes could protect networks used today.

Troll Hunters: The Twitterbots That Fight Against Online Abuse
New Scientist (08/03/16) Sally Adee

Internet companies have made promises to deal with online abuse, but their efforts using human moderation have come up short. As a result, more researchers are looking to use bots as a solution to curbing online harassment. The simplest way bots can help is with block lists, but the approach only works if someone adds an abusive account to the block list. Some researchers are working to automate the detection of harassment, but the Illinois Institute of Technology's Libby Hemphill says humans do not agree on what constitutes harassment, so that makes it difficult to train computers. Argue-bots have been developed to distract trolls from their human targets by drawing their attention and engaging with them. New York University's Kevin Munger developed a bot in an attempt to manipulate a troll's sense of group dynamics online. However, he does not want bots to be the endgame. "I don't envision an army of bots telling people to behave themselves," Munger says.

Flexible Wearable Electronic Skin Patch Offers New Way to Monitor Alcohol Levels
UC San Diego News Center (08/02/16) Liezel Labios

University of California, San Diego (UCSD) researchers have developed a sensor that can accurately monitor alcohol level within 15 minutes. The sensor is designed to measure a person's blood alcohol level from sweat and wirelessly transmit the data to a laptop, smartphone, or other mobile device. The researchers note the device is flexible, can be worn on the skin, and could be used by doctors and police officers for continuous, non-invasive, and real-time monitoring of blood alcohol content. The technology consists of a temporary "tattoo" that sticks to the skin, induces sweat, and electrochemically detects the alcohol level. The device also includes a portable flexible electronic circuit board connected to the tattoo by a magnet, which can communicate the information to a mobile device via Bluetooth. Researchers tested the alcohol sensor on nine healthy volunteers who wore the tattoo on their arms before and after consuming an alcoholic beverage, and found the readouts accurately reflected the wearers' blood alcohol concentrations. "What's also innovative about this technology is that the wearer doesn't need to be exercising or sweating already," notes UCSD professor Patrick Mercier. "The user can put on the patch and within a few minutes get a reading that's well correlated to his or her blood alcohol concentration."

UNH Lab Peers Under the Hood as Software Invades Network Systems
Concord Monitor (NH) (08/02/16) David Brooks

The University of New Hampshire InterOperability Lab (IOL) on Monday shifted its focus from networking hardware to software-defined networking (SDN). IOL senior executive Timothy Winters describes SDN as "the next evolution in networking; it's a dynamic change. It's going from human interaction with having your network be programmed." With an operating budget of approximately $10 million underwritten by industry partners, Winters says the IOL is staffed by about 100 paid undergraduates, along with about 10 graduate students and 40 full-time personnel. The lab has hosted several events, known as AppFest and PlugFest, oriented around non-proprietary systems developed and promoted by the Open Networking Foundation. The success of those events and the sustained interest encouraged the IOL's expansion into SDN. Winters also says the migration to SDN aids the lab's role as an educational resource. "We want to get students working with the technology because it is what is going on out in the world," he notes. "We want to give them real-world experience."

UW Lab Creates Computer Programs to Self-Teach
Laramie Boomerang Online (08/01/16) Thaddeus Mast

The University of Wyoming's (UW) Evolving Artificial Intelligence Laboratory, founded by professor Jeff Clune, focuses on networks' internal functioning. Clune says, "once a network has learned to do something, like classify deer and giraffes, you can try and understand how it's doing that. We call it 'AI neuroscience,' because it's actually very similar to what neuroscientists do." One program developed at the lab is designed to identify wildlife in photos from camera traps in Africa, using a dataset of unlabeled images to train on so it can learn on its own, says postdoctoral researcher Arash Norouzzadeh. The program analyzed 50,000 wildlife photos in UW's supercomputer over five days, and it can correctly identify an animal more than 90 percent of the time. Clune has submitted a grant application to the U.S. National Science Foundation to fund the project, and he says his team will develop it further while waiting for approval. The lab also supports the projects of five other postdoctoral researchers, including finding and modulating specific parts of a program, which has potential for enhancing the speed and simplicity of creating complex robotics programs. Clune stresses the need for AI students to actively work with the robots in his lab. "It's always more difficult, because working with hardware takes more time, but it shouldn't actually be too hard," he says.

Encryption's Quantum Leap: The Race to Stop the Hackers of Tomorrow
ZDNet (08/02/16) Steve Ranger

Researchers are looking into the construction of new quantum-proof cryptography in order to thwart quantum-based schemes that future hackers could potentially use to crack sensitive data. "If large-scale quantum computers are ever built, they will be able to break many of the public-key cryptosystems currently in use," warns the U.S. National Institute of Standards and Technology (NIST). "This would seriously compromise the confidentiality and integrity of digital communications on the Internet and elsewhere." NIST is requesting comments on a new process to find and assess public-key cryptographic algorithms that quantum computers cannot decrypt. NIST's goal is to create systems that are resistant to both quantum and classical computers, as well as interoperable with existing communications protocols and networks. The agency is investigating preliminary evaluation criteria for quantum-resistant public-key cryptography standards, which is slated for finalization by year's end. NIST then will start accepting proposals for such encryption, digital signatures, and key exchange algorithms, with a deadline in late 2017, followed by three to five years of public scrutiny before their acceptance as standards.

Smartwatch Interface Could Improve Communication, Help Prevent Falls at Nursing Homes
Binghamton University (08/01/16)

Researchers at Binghamton University are developing a smartwatch application to improve communication and notification systems for nursing homes. The proposed design integrates all of the existing safety systems at nursing homes, such as call lights, chair and bed alarms, wander guards, and calling-for-help functions. "With our system, we provide an informative and customized message for different alarms," says Binghamton professor Huiyang Li. The research team believes the app could help certified nursing assistants (CNAs) respond to alerts more quickly and help prevent falls. Every CNA who uses the app will see a different display, as it will be personalized to the user's specific task assignment, the researchers say. When a resident triggers an alert, a message will pop up on everyone's screen indicating who the resident is, their room number, and the type of alert, such as an exit from a chair. The team notes the final design was well received by nursing experts in geriatric care and the system reduces staff response time to alarms. "The improvement of notification will potentially help staff to do a better job and, eventually, improve patient safety," Li says.

New Silicon Structures Could Make Better Biointerfaces
Argonne National Laboratory (08/01/16) Kate Thackrey

A team of researchers has engineered silicon particles one-50th the width of a human hair, which could lead to "biointerface" systems designed to make nerve cells fire and heart cells beat. The researchers note the particles can establish unique biointerfaces on cell membranes, because they are deformable but can still yield a local electrical effect. The team from the University of Chicago, Northwestern University, the University of Illinois at Chicago, and the U.S. Department of Energy's Argonne National Laboratory created the mesostructured silicon, named for its complex internal structures of nanoscopic wires, using a process called nano-casting. To make the particles, each between one and five micrometers in size, researchers filled the beehive structure of synthetic silicon dioxide with semiconductive silicon the same way a blacksmith would pour molten metal into a cast-iron mold. The outer mold was then etched away with acid, leaving behind a bundle of wires connected by thin bridges. The material has many potential applications in biomedicine, because the particles and light can be used to excite many types of cells. To test whether the particles could change the behavior of cells, the team injected a sample onto cultured rat dorsal root ganglia neurons. They successfully activated the neurons using pulses of light to heat the silicon particles, causing current to flow through the cells.

Of Heartbeats, Bones and Brushstrokes
Duke Today (08/01/16) Robin A. Smith

A five-year, $1.5-million award from the Simons Foundation will enable Duke University professor Ingrid Daubechies to expand her collaborative research involving mathematics and electrical engineering. Daubechies will use the award to build machine learning tools that offer new ways to mine data. Some researchers have argued new approaches are needed in order to keep up with the deluge of data. "The large volumes of data that are being generated left and right definitely present a serious challenge," Daubechies says. Daubechies has been focusing on training computers to churn through electrocardiogram tracings, high-resolution scans of fossils, paintings, and other complex digital data and work things out automatically. One new technique Daubechies is working on involves comparing three-dimensional (3D) shapes, which relies on computational geometry to make the process less time-consuming and subjective for computers. For the past seven years, Daubechies has been working with fossil experts to analyze 3D images of bones and teeth.

Abstract News © Copyright 2016 INFORMATION, INC.
Powered by Information, Inc.

To submit feedback about ACM TechNews, contact: [email protected]