Welcome to the January 20, 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).
HEADLINES AT A GLANCE
Shrinking the Haystack
Technologists are assisting counter-terrorist forces in the West and abroad in a major way by producing software that can show which areas should be searched for hideouts and weapons caches or be put under surveillance. They are modifying existing mapping software to produce "geographic profiling" programs. Counter-terrorist forces feed the programs the times and coordinates of bombings and the location of extremist groups' leafleting and graffiti. The programs then crunch the data, and analyze it in the context of information about the country's terrain, road network, ethnic make-up, and shifting patterns of tribal alliance. The U.S. Army developed SCARE-S2 geoprofiling software for use in Afghanistan, and more than 90 intelligence agencies around the world use Rigel Analyst from ECRI in Vancouver. The ArcGIS software analyses global-positioning system data provided by smartphones and other gadgets that are equipped with global-positioning system kits. The number of pertinent actions that can be plotted is booming, so the software is likely to become an even more critical tool in years to come. The technology works particularly well in areas in which patterns of belief are tied closely to geography, notes Brent Smith at the University of Arkansas' Center for Advanced Spatial Technologies.
'Hack the Dinos' Helps Paleontologists
The New York Times (01/18/16) Kenneth Chang
The American Museum of Natural History's Hack the Dinos challenge in November sought to encourage coders, graphic designers, and data scientists to create software to aid paleontologists in their research. Participants were challenged to fulfill a digital tool wish list, which included automated skull analysis, mobile field notebooks, and digital conversion of cladograms. One of the teams produced a smartphone app called Foss'l to help ease the workload of museum staffer Carl Mehling, who is responsible for answering fossil queries. "We wanted to combine the experience of being able to submit your fossils to poor Carl with a learning experience," says Parsons School of Design student Jessie Contour. She describes the app as "Tinder for fossils." Postdoctoral paleontological researcher Jack Tseng was particularly impressed by an app called Sapling Detector, which converts paper evolutionary trees into an electronic format. "That's something I would use right away," Tseng says. "It solves the problem we were trying to solve."
Open Source GPU Could Push Computing Power to the Next Level
Binghamton University (01/19/16)
Binghamton University researchers say they have achieved a landmark by being the first to use an open source graphics processing unit (GPU) for research. They have developed a synthesizable GPU architectural model called Nyami for general-purpose and graphics-specific workloads. "While simulators may take shortcuts, an actual synthesizable open source processor can't cut any corners, so we can say that any experimental results we get are especially reliable," says Binghamton professor Timothy Miller. He notes the results of their work will assist other researchers in the development of their own GPUs and raise computing power to the next level. Nyami is the first GPU that can be modified by scientists and hobbyists to get a sense of how changes may impact mainstream chips. Roku's Jeff Bush, who co-developed Nyami, says GPU makers' closely held secrecy about their chip specifications "prevented open source developers from writing software that could utilize that hardware. [Nyami] makes it easier for other researchers to conduct experiments of their own, because they don't have to reinvent the wheel." The researchers think these findings could ease the use of processors by the research community, as well as the exploration of different design tradeoffs.
Drones Dodge Obstacles
MIT News (01/19/16) Adam Conner-Simons
Researchers from the Massachusetts Institute of Technology's (MIT) Computer Science and Artificial Intelligence Laboratory (CSAIL) announced two projects this week that demonstrated software enabling drones to make precise stops so they can execute hairpin movements over, under, and around 26 obstacles in a simulated "forest." The Iterative Regional Inflation by semidefinite programming algorithm and a mixed-integer semidefinite program were combined to fragment space into "obstacle-free regions" and then connect them to find a single collision-free route. "Using free-space segments is a more 'glass-half-full' approach that works far better for drones in small, cluttered spaces," says researcher Benoit Landry. The second CSAIL project involved a fixed-wing plane that avoids obstacles without any advanced knowledge of the space, even in the presence of wind gusts and other dynamics. The researchers were able to enable the plane to avoid obstacles by pre-programming a library of 40 to 50 "funnels" representing the worst-case behavior of the system, calculated by a thorough verification algorithm. "As the drone flies, it continuously searches through the library to stitch together a series of paths that are computationally guaranteed to avoid obstacles," notes MIT Ph.D. student Anirudha Majumdar. "Many of the individual funnels will not be collision-free, but with a large-enough library you can be certain that your route will be clear."
Tweets and Reddit Posts Give Snapshot of Our Changing Language
New Scientist (01/18/16) Hal Hodson
Lancaster University researchers have developed software that charts the rise of language online. The software established methods lexicographers use to chart the popularity of words, translated them into algorithms, and applied them to 22 million words drawn from Twitter and Reddit posts. The researchers want to examine the niche portions of the Internet and chart novel language making its way into the mainstream. "If we see an innovation taking off on Reddit or Twitter, the question is what point is it going to appear in a newspaper," says Lancaster University researcher Matthew Rowe. The algorithm picks words that have gone through a sudden rise in popularity. The Oxford University Press is using automated methods similar to the Lancaster University technique to monitor language evolution. In December, Oxford Dictionaries added 500 new words based on this method. Online communication is not necessarily accelerating language evolution, but rather making that evolution easier to see, according to Katherine Martin, head of U.S. Dictionaries at Oxford University Press. In addition, she notes the Internet provides a way to distribute new language around the world that was previously limited to major TV stations.
How Future Cars Will Predict Your Driving Maneuvers Before You Make Them
Technology Review (01/15/16)
Cornell University researcher Ashesh Jain and colleagues have developed a prototype car safety system that can predict a driver's next maneuver about three seconds ahead of time, which can be applied to the identification and prevention of potential collisions. The team employed artificial intelligence algorithms to analyze multiple data streams and spot signals of an impending maneuver. They first trained their machine with a large data set compiled by cameras watching both the driver and the road ahead, paired with global-positioning system data, street maps, and a speed logger. The team then culled data from 10 motorists who spent two months compiling more than 1,000 miles of freeway and city driving, and this data was manually annotated to identify 700 maneuvers or "events." The data was fed to several number-crunching programs to mark the conditions under which a driver would turn left or right, change lanes to the left of right, or keep driving in a straight line. The best-performing algorithm correctly guessed a future maneuver about 90 percent of the time, and on average it was able to make its prediction 3.5 seconds before the maneuver actually happened. The algorithm still needs to be tested under hazardous driving conditions, and it remains uncertain how automakers can use the collected information to prevent accidents.
App Aims to Make Cultural Heritage Interesting and Interactive
CORDIS News (01/15/16)
A mobile application designed to increase customer engagement with cultural heritage sites is now available from the European Union-funded TAG CLOUD project. The COOLTURA app uses mobile features such as geo-location and real-time interaction to provide users with a more personalized cultural experience. Users access and use cultural content from the cloud-based COOLTURA platform. The app is available for smartphones and tablets, and can be integrated with wearable devices, such as the Apple Watch. "TAG CLOUD uses the cloud to provide an adaptive cultural experience according to user interests, likes, and habits," says project coordinator Maria Fernanda Cabrera-Umpierrez from the Technical University of Madrid in Spain. She says the COOLTURA concept was successfully deployed at three sites, including the monument complex of the Alhambra, and the TAG CLOUD team expects to roll it out to other cultural institutions. The TAG CLOUD team says another app for discovering, creating, and sharing digital stories related to places, via social media, called Stedr, is not fully ready for market. The project concludes this month.
Hacking Team's Leak Helped Researchers Hunt Down a Zero-Day
Wired (01/13/16) Kim Zetter
Researchers at Kaspersky Lab have for the first time discovered a valuable zero-day exploit after intentionally going on the hunt for it. They found a remote-code execution exploit that attacks a vulnerability in Microsoft's Silverlight software. Microsoft called the vulnerability "critical" in a patch recently released to customers. The vulnerability enables an attacker to infect a system after getting the user to visit a malicious website where the exploit resides--usually through a phishing email that tricks the user into clicking on a malicious link. Kaspersky Lab caught the Silverlight exploit in late November after the zero-day infected a customer's machine, but it took a clever lure and months of patiently waiting. The first clue came from hacked Hacking Team emails discussing Silverlight. "This is actually the first time that we are succeeding in catching something that we planned on hunting," says Costin Raiu, head of Kaspersky's Global Research and Analysis Team. "It was probably a bit of intuition and luck. If the compiler would have removed these [debugging] strings, then obviously [there would have been] no luck for me."
Sensors Slip Into the Brain, Then Dissolve When the Job Is Done
IEEE Spectrum (01/18/16) Tekla S. Perry
University of Illinois researchers led by professor John Rogers have developed flexible sensors that can operate accurately inside the human body for at least five days before dissolving. The pressure and temperature sensors are intended to be implanted directly into the brain, and will completely dissolve within a few weeks. The researchers demonstrated the devices in rats, using soluble wires to transmit the signals, as well as a wireless version. The technology can be adapted to sense fluid flow, motion, pH, and other parameters, and could be implanted in the heart, other organs, or in the skin. The devices are microelectromechanical systems made out of a membrane of polylactic-co-glycolic acid, a biodegradable polymer common in medical applications. The membrane sits on a substrate of nanoporous silicon or a metal foil, which is etched with trenches that create an air cavity, enabling the membrane to deflect in response to pressure changes in the surrounding fluid. The researchers plan to continue to improve the technology, pushing its useful life to four weeks before the devices dissolve. The next stage of the research involves resorbable devices that go beyond sensing to actively helping with treatment.
Malicious Coders Will Lose Anonymity as Identity-Finding Research Matures
U.S. Army Research Laboratory (01/15/16) Joyce P. Brayboy
Researchers from three universities and the U.S. Army Research Laboratory (ARL) are trying to address the problem of identifying authors of malicious code and software. At the recent Chaos Computer Congress in Hamburg, Germany, Princeton University researcher Aylin Caliskan-Islam presented a code stylometry study, which examined samples from 1,600 coders. With 94-percent accuracy, a machine-learning algorithm could determine the author of a particular code excerpt. The team of researchers also examined executable binary authorship using a novel set of features, such as decompiling the executable binary to source code, according to their recent paper. Along with Caliskan-Islam, the team includes Fabian Yamaguchi from the University of Gottingen and Edwin Dauber from Drexel University. "Attribution is a real challenge [as opposed to detection], as it is done manually by experts who have to reconcile forensics following an attack," says Richard Harang, ARL network security researcher and technical lead. "We are developing a toolkit to make it a lot faster and cheaper to support analysts in identifying bad actors." A limitation is success depends on having existing samples from potential authors and malware authors' ability to mask software.
Scientists Demonstrate Basics of Nucleic Acid Computing Inside Cells
Georgia Tech News Center (01/19/16) John Toon
Scientists have demonstrated basic computing operations inside a living mammalian cell, technology that could lead to an artificial sensing system that could control the behavior of cells. The researchers used DNA strand displacement, a technology that has been widely used outside of cells for the design of molecular circuits, motors, and sensors. The researchers modified the process to provide both AND and OR logic gates able to operate inside the living cells and interact with native messenger RNA. Strand displacement reactions are the biological equivalent of the switches or gates that form the foundation for silicon-based computing. "The whole idea is to be able to take the logic that is used in computers and port that logic into cells themselves," says Georgia Institute of Technology/Emory University professor Philip Santangelo. The research could lead to the development of devices that could, for example, sense an aberrant RNA and then shut down cellular translation or induce cell death. The next step is to use the switching of nucleic acid computers to trigger the production of signaling chemicals, which would prompt the desired reaction from the cells. The research was supported by the U.S. Defense Advanced Research Projects Agency and the U.S. National Science Foundation.
UCI News (01/14/16) Brian Bell
University of California, Irvine professor Sharad Mehrotra and colleagues are transforming Donald Bren Hall into a testbed for safeguarding people's online identities and confidential data. The team is installing sensors, scanners, and surveillance cameras to test the limits of privacy for humans interacting with smart buildings. In a hypothetical scenario, the building's surveillance camera would capture a person's image, a wireless network would "sniff" the mobile phone in their backpack, and its management system would know which office the person enters, the computer they select, and whether that room's light is on. Mehrotra wants to develop novel technologies to enable building occupants to state their preferences. "We're trying to study where there's a trade-off between what you're willing to share and what you get as a utility," he says. The U.S. Defense Advanced Research Projects Agency has provided a $5-million grant for the research project. This four-year, multi-institutional effort will attempt to answer concerns about the status of personal information in the cloud environment, mobile computing, and the Internet of Things. After installing sensors throughout the hall, the team will develop applications for monitoring energy usage and human presence in workplaces.
Scientists, Educators Present Research on All Aspects of How We Learn
The Hub (01/13/16)
Last week's biennial Science of Learning Symposium and fourth annual Symposium on Excellence in Teaching and Learning in the Sciences at Johns Hopkins University (JHU) convened many scientists and educators to showcase the latest human learning research. An interdisciplinary approach to understanding lifelong learning was emphasized by JHU provost Robert C. Lieberman, who noted the symposia were designed to nurture deep and ongoing collaboration. Deep neural networks were JHU professor Alan Yuille's topic of discussion, with particular focus on image-recognition software. Yuille said although computers can be taught to recognize certain shapes within an image and automatically produce captions, they can be fooled relatively easily by adversarial testing, which also can improve the software. Meanwhile, professor Jason Eisner discussed foreign language learning via assistive artificial intelligence, and argued designing software to learn is a better approach than designing it with all of its knowledge already coded in. Among the presentations highlighted at the symposia were those concerning neurological activity in the brain during learning or training. Advanced imaging methods enable researchers to view learning as it occurs in live test subjects, while presenters also discussed linguistics and the way language and reading skills are learned in children and adults.
Abstract News © Copyright 2016 INFORMATION, INC.
To submit feedback about ACM TechNews, contact: [email protected]
Current ACM Members: Unsubscribe/Change your email subscription by logging in at myACM.