Association for Computing Machinery
Welcome to the April 16, 2012 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).

HEADLINES AT A GLANCE


Designing the Interplanetary Web
European Space Agency (04/13/12)

The European Space Agency (ESA) is pursuing projects such as providing reliable Internet access on the Moon, and controlling a planetary rover from a spacecraft in deep space. Observation or navigation satellites orbiting Earth and astronauts transmitting images in real time from the International Space Station need to send data back to Earth. The complexity of sharing information across space is set to grow, and such activities will need to be interconnected, networked, and managed. "We are researching how today's technical standards for devices like mobile phones, laptops, and portable computers can be applied to a new generation of networked space hardware," says ESA's Nestor Peccia. "But our future focus goes well beyond just networking; we're looking at how agencies like ESA and [the U.S. National Aeronautics and Space Administration] cooperate in orbit and how to interchange data in real time between different organizations' spacecraft and ground stations, as well as reliable technical standards for spacecraft navigation and flight control." In October, ESA will simulate orbiter-rover communication links at a planet like Mars.


Researchers Tune In to the Internet Buzz
Wall Street Journal (04/13/12) Melinda Beck

University of Pennsylvania researchers recently launched the Mining Internet Messages for Evidence of Herbal-Associated Adverse Events (MICE) project, which involves mining message boards and Twitter feeds to see what breast and prostate cancer patients are saying about herbal and nutritional supplements as treatments. Even if there is no scientific evidence to support what people post, it is useful to identify areas that would merit further study, says University of Pennsylvania researcher John Holmes. The Internet can be a great source for information epidemiology and "we can learn a lot about public sentiment, public attitudes, and public knowledge," agrees University of Toronto professor Gunther Eysenbach. However, analyzing Web conversations raises some ethical and privacy issues. MICE researchers only mine discussion sites that require participants to register and explicitly state in their terms of use that any information posted will become public. The researchers also developed an anonymizer program that scrubs out names, locations, or other identifiers. Each person who posts is given a unique identifying code so that researchers can follow their conversation threads and understand what they mean.


Artificial Intelligence Could Be on Brink of Passing Turing Test
Wired News (04/12/12) Brandon Keim

The vast amounts of raw data and new sophisticated techniques for collecting, organizing, and processing that data are revolutionary advances in information technology that could lead to the solving of the Turing test, says French National Center for Scientific Research scientist Robert French. Probabilistic and connectionist approaches are utilized by many of today's real-world artificial intelligence technologies, including autonomous cars, Google searches, and automated language translation. A machine that can pass the Turing test would be similar to being able to record and access every word a user has spoken, heard, written, or read, as well as all the visual, auditory, tactile, and olfactory senses experienced over the course of a lifetime, according to French. The software also would be able to catalog, analyze, correlate, and cross-link everything in that set of data. University of Michigan's Satinder Singh says big data could be the source of building a flexibly competent intelligence, but he cautions that there are many questions that haven’t been studied. “In order to be broadly and flexibly competent, one needs to have motivations and curiosities and drives, and figure out what is important,” Singh says. “These are huge challenges.”


Novel Coding Technique Patented by A*STAR Researchers
A*STAR Research (04/11/12)

A*STAR Data Storage Institute researchers have developed an algorithm for correcting errors that occur when information is stored and read out incorrectly. The algorithm significantly enhances the error tolerance of spin-torque transfer magnetic random access memory (STT-MRAM), which is a key contender for the future of non-volatile memories. "This is a breakthrough work that will help provide bigger tolerances and ease the engineering challenges in STT-MRAM material and process development," says A*STAR researcher Pantelis Alexopoulos. The researchers also developed a new design of the memory sensing and detection architecture that is based on soft decision decoding. They say the technique leads to significantly fewer decoding errors than hard decision decoding. The design also features a soft-output channel detector, which measures the probabilities of the bits read out being set as 0 or 1, and feeds the information into the soft decision decoding process. The researchers have shown that the new design achieves a 20 percent increase in the tolerance towards variations in the electrical resistance of devices.


Data Mining Opens the Door to Predictive Neuroscience
Ecole Polytechnique Federale de Lausanne (04/11/12) Lionel Pousaz

Ecole Polytechnique Federale de Lausanne (EPFL) researchers have discovered rules that relate the genes that a neuron switches on and off to several features of the neuron itself. The discovery increases the likelihood that it will be possible to predict much of the fundamental structure and function of the brain without having to measure every aspect of it. "It is the door that opens to a world of predictive biology," says EPFL's Henry Markram. The researchers used a dataset that included the expression of 26 genes encoding ion channels in different neuronal types from a rat brain. The researchers also used data classifying those types according to various aspects of the neuron itself. The researchers found that, based on the classification data alone, they could predict those previously measured on ion channel patterns with 78 percent accuracy. "This shows that it is possible to mine rules from a subset of data and use them to complete the dataset informatically," says EPFL's Felix Schurmann. Researchers could use the rules to study the different genes in regulating transcription processes. This discovery could lead to a new era of predictive biology and accelerate progress toward understanding and modeling the human brain.


Physicists Create First Long-Distance Quantum Link
ScienceNOW (04/11/12) Jim Heirbaut

Max Planck Institute of Quantum Optics researchers say they have built the first true quantum link using two widely separated atoms, which could lead to a complete network if many such links are combined. The researchers say they successfully entangled two atoms in separate labs on opposite sides of the street. Two atoms are entangled when they are both in an uncertain two-ways-at-once state, but their states are perfectly correlated. The researchers placed each atom between two highly reflective mirrors 0.5 millimeters apart, creating an optical cavity. By applying an external laser to atom A, the researchers caused a photon emitted by that atom to escape from its cavity and travel through a 60-meter-long optical fiber to the cavity across the street. When the photon was absorbed by atom B, the original quantum information from the first atom was transferred to the second. The entanglement could in principle be extended to a third atom, which makes the system scalable to more than two nodes, according to the researchers. Although this is a very important advance, "there is still a great deal of work to be done before the technology is practical," notes Toshiba researcher Andrew Shields.


Building a Smarter Forest
The Atlantic (04/11/12) Jared Keller

B.S. Abdur Rahman University computer scientists M.P. Sivaram Kumar and S. Rajasekaran recently published a research paper that says the existing methods for preventing and extinguishing forest fires are inefficient. The researchers argue that a robotic system deployed throughout wooded areas would increase firefighters' ability to identify emerging threats and would more effectively plot the path of a wildfire. The researchers describe a grid-based system of automated drones, designed to detect abnormal changes in temperature and relay data back to a command center. "The automatic forest fire detection and extinguishing system consists of nodes deployed deterministically in a forest area and all the nodes know their location based on coordinate values of a matrix," the paper says. The researchers also say that on-the-ground nodes are much better than existing satellite detection systems that detect fires relatively late. In addition, they note that U.S. federal agencies and wireless carriers are planning to deploy the Wireless Emergency Alerts (WEA) system, which will notify people of severe weather that poses imminent threats to their safety. When added to the WEA system, the researchers' robotic system could provide people with enough advance warning to protect them from natural disasters.


Face Recognition Could Catch Bad Avatars
New Scientist (04/11/12) Jacob Aron

University of Louisville researchers are developing the field of artificial biometrics, known as artimetrics, to serve as a way to authenticate and identify non-biological agents such as avatars, physical robots, and chatbots. The researchers, led by Roman Yampolskiy, have developed facial recognition techniques specifically designed for avatars. "Not all avatars are human looking, and even with those that are humanoid there is a huge diversity of color," Yampolskiy says. Therefore, the software uses a large variety of colors to improve the recognition of avatars. The researchers also are studying how to match a human face to an avatar generated from that face. Combining the color-based technique with existing facial recognition software led to the best results, suggesting that it could be possible to track users between the physical and virtual worlds. Yampolskiy also intends to create recognition algorithms for robots, forecasting that, in Japan at least, autonomous robots may become sufficiently ubiquitous to require their own identification. Meanwhile, Yampolskiy is working with his Louisville colleagues to develop methods for determining the legitimacy of chatbots. The researchers fed text written by chatbots into software originally designed to identify human authors and found that they were often able to determine the chatbot responsible.


Toward a Modular Defense Against Hackers
Lehigh University (04/10/12) Kurt Pfitzer

Lehigh University professor Gang Tan has developed automated techniques to scan for errors in large software systems. Tan and Lehigh researchers also recently received a five-year CAREER Award from the U.S. National Science Foundation to study and develop modular software that is less vulnerable to system-wide attacks by hackers. The researchers want to apply the principle of least privilege to software systems. "The principle of least privilege is like the separation of powers in a political system," Tan notes. He says the researchers have made progress in privilege separation in software environments, but challenges remain with operating system portability, high runtime overhead, architectural flexibility, and compositional reasoning. “These new tools and methodologies will make the principle of least principle easier to apply to big software systems," Tan says. "By monitoring information flow at the binary instead of the source-language level, it will be easier to check the security properties of individual modules, prevent malicious information flow between modules, and allow only benign information flow."


Web-Based Tool Produces Fast, Accurate Autism Diagnosis
Harvard University (04/10/12) Katie DuBoff

Harvard Medical School researchers have developed algorithms and associated deployment mechanisms to rapidly and accurately detect autism. The algorithms are designed to work within a mobile architecture, combining a set of questions and a short home video of the subject, to enable rapid online assessments, which could reduce the time for autism diagnosis by nearly 95 percent. "We believe this approach will make it possible for more children to be accurately diagnosed during the early critical period when behavioral therapies are most effective," says Harvard professor Dennis Wall. When children are evaluated for autism, they typically take the Autism Diagnostic Interview, Revised (ADI-R) or the Autism Diagnostic Observation Schedule (ADOS) exam, which measures several behaviors in children. Wall used machine-learning techniques to study the results of the ADI-R from the Autism Genetic Research Exchange for more than 800 individuals diagnosed with autism to find redundancies across the exam. The researchers found that only seven questions were sufficient to diagnose autism with nearly 100 percent accuracy. The researchers applied similar techniques to the ADOS exam, and classified more than 1,050 individuals with near perfect sensitivity and slightly less than 95 percent specificity.


Cooperating Mini-Brains Show How Intelligence Evolved
Live Science (04/10/12) Stephanie Pappas

Trinity College Dublin researchers recently developed computer simulation experiments to determine how human brains evolved intelligence. The researchers created artificial neural networks to serve as mini-brains. The networks were given challenging cooperative tasks and the brains were forced to work together, evolving the virtual equivalent of increased brainpower over generations. "It is the transition to a cooperative group that can lead to maximum selection for intelligence," says Trinity's Luke McNally. The neural networks were programmed to evolve, producing random mutations that can introduce extra nodes into the network. The researchers assigned two games for the networks to play, one that tests how temptation can affect group goals, and one that tests how teamwork can benefit the group. The researchers then created 10 experiments in which 50,000 generations of neural networks played the games. Intelligence was measured by the number of nodes added in each network as the players evolved over time. The researchers found that the networks evolved strategies similar to those seen when humans play the games with other humans. "What this indicates is that in species ancestral to humans, it could have been the transition to more cooperative societies that drove the evolution of our brains," McNally says.


Humanoid Robot Has Muscles, Joints and Tendons
Mashable (04/06/12) Joann Pan

Scientists at the universities of Sussex, Zurich, and Belgrade have built Eccerobot, a robot with muscles, bones, and tendons based on the mechanics of the human body. The team believes the mechanics inspired by the human build will enable Eccerobot to walk and move its arms with more speed and rhythm than other robots. "You can use the passive compliance to make it absorb the energy in the right way to allow for safe interaction and to store energy in the muscles to produce fast movement," says Zurich's Hugo Gravato Marques. The researchers completed a half-torso model that sits on a mobile platform. The arms have numerous parts that detect strain, and the bones are made from a thermoplastic material that morphs into shape with heat. The robot can hold objects, shake hands, and smoothly lift its arms. High-speed, high-definition cameras serve as eyes, an audio detection system acts as its ears, and force-sensitive-resistors on its fingertips and palms give Eccerobot a sense of touch.


Abstract News © Copyright 2012 INFORMATION, INC.
Powered by Information, Inc.


To submit feedback about ACM TechNews, contact: [email protected]
Current ACM Members: Unsubscribe/Change your email subscription by logging in at myACM.
Non-Members: Unsubscribe