Association for Computing Machinery
Welcome to the August 16, 2013 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).


NSA Broke Privacy Rules Thousands of Times Per Year, Audit Finds
The Washington Post (08/15/13) Barton Gellman

The U.S. National Security Agency (NSA) has violated privacy rules or overreached its authority thousands of times annually since 2008, primarily through unauthorized surveillance of American and foreign intelligence targets in the U.S., according to an internal audit and other documents. The infractions include major violations of law as well as typographical errors that led to misdirected interception of U.S. emails and telephone calls. One of the documents, provided to The Washington Post by former NSA contractor Edward Snowden earlier this summer, instructs NSA workers to eliminate details in reports to the Justice Department and the Office of the Director of National Intelligence. The May 2012 internal audit found 2,776 incidents in the preceding 12 months of unauthorized collection, storage, access to, or distribution of legally protected communications. "We're a human-run agency operating in a complex environment with a number of different regulatory regimes, so at times we find ourselves on the wrong side of the line," says a senior NSA official. "You can look at it as a percentage of our total activity that occurs each day. You look at a number in absolute terms that looks big, and when you look at it in relative terms, it looks a little different."
View Full Article - May Require Free Registration | Return to Headlines | Share Facebook  LinkedIn  Twitter 

Google Is Preparing for Screenless Computers
Quartz (08/15/13) Christopher Mims

Google aims to fulfill its vision of a future in which people's primary interaction with computers is vocal by enabling voice commands in all of its products, and making such engagement as seamless, simple, and intuitive as current keyboard and touchscreen technology. "What we're really trying to do is enable a new kind of interaction with Google where it's more like how you interact with a normal person," says Scott Huffman, head of Google's Conversation Search group. He says that realizing the level of simplicity that the group is aiming for requires leveraging Google's complete body of accumulated knowledge about the real world. The primary tool for this is the Google Knowledge Graph, a database of people, places, and things. Huffman says Google seeks to apply its intimate knowledge of its users to make automatic speech recognition better than that of humans. The Google voice interface already can comprehend pronouns like he, she, and it, but among the challenges of achieving voice-only human-computer interaction is imbuing voice control with reliability and error correction. A second challenge is the need for people to learn an entirely new computer control methodology, as well as the capabilities of the software that might be employed in this manner.

IBM, Universities Partner on 'Big Data' Skills Training
IDG News Service (08/14/13) Chris Kanaracus

IBM recently announced nine partnerships with global universities to help prepare the next generation of big data professionals. The new offerings will include a master of science in business analytics at George Washington University, an undergraduate course on big data analytics at the University of Missouri, and a center for business analytics at the National University of Singapore. Participants will have access to IBM big data and analytics software solutions, case studies, and guest lecture appearances by IBM thought leaders, says an IBM spokesperson. The schools will benefit from access to technology and other resources, and IBM and the industry will benefit by grooming the next-generation of big data consultants, data scientists, and developers. Over the next eight years, demand for data analytics skills will grow 24 percent, according to the U.S. Bureau of Labor. Big data jobs have increased 127 percent on tech jobs board, with about 1,500 daily job listings, making it the site's fastest-growing job category, while other major categories have stayed mostly flat year-over-year.

Encryption Is Less Secure Than We Thought
MIT News (08/14/13) Larry Hardesty

Researchers at the Massachusetts Institute of Technology and the National University of Ireland recently published a paper showing that most information-theoretic analyses of cryptographic systems have been based on an incorrect mathematical assumption, making cryptography less secure than previously believed. The researchers found that information-theoretic analyses of secure systems have generally relied on Shannon entropy, which was written in 1948 and is based on the average probability that a particular string of bits will occur in a certain type of digital file. However, cryptography is not concerned with the average case, but rather with the worst-case scenario. To gain additional information, codebreakers only need one reliable correlation between encrypted and unencrypted versions of a file. Other theories of entropy have since emerged that apply more weight to improbable outcomes, and thus provide more accuracy when applied to codebreaking. However, the findings do not mean that current cryptographic systems are fundamentally insecure, says Georgia Institute of Technology professor Matthieu Bloch. "My guess is that it will show that some of them are slightly less secure than we had hoped, but usually in the process, we'll also figure out a way of patching them," Bloch says.

Researchers Building a Computer Chip Based on the Human Brain
Boise Update (08/14/13) Kathleen Tuck

Boise State University researchers are developing a chip with a computing architecture resembling the human brain. "By mimicking the brain's billions of interconnections and pattern recognition capabilities, we may ultimately introduce a new paradigm in speed and power, and potentially enable systems that include the ability to learn, adapt, and respond to their environment," says principal investigator Elisa Barney Smith. The key focus of the project is a memristor, or a resistor that can be programmed to a new resistance by the application of electrical pulses. The team's research expands upon recent work from researchers who have derived mathematical algorithms to explain the electrical interaction between brain synapses and neurons. "By employing these models in combination with a new device technology that exhibits similar electrical response to the neural synapses, we will design entirely new computing chips that mimic how the brain processes information," Barney Smith says. The chips will be able to consume power at an order of magnitude lower than current computing processors, despite the fact that they match existing chips in physical dimensions. After the team builds an artificial neural network, it will seek to engage neurobiologists in parallel to what they are doing now.

Researchers Develop Acoustic Based Data Transfer System for Phones
Computerworld Australia (08/14/13) Rohan Pearce

Microsoft researchers have developed Dhwani, a system that offers near-field communications (NFC)-like capabilities without requiring dedicated wireless hardware. Dhwani uses sound, employing the speaker and microphone on phones to securely exchange data at speeds of up to 2.4 Kbps, which is "sufficient for most existing NFC applications," the researchers note. Although the range of Dhwani is significantly shorter than Wi-Fi or Bluetooth, it is similar to NFC, and will work effectively over a distance of a few centimeters. Dhwani is software-based and therefore does not require specialized hardware for transfers, and the researchers also say the technology is more secure than NFC. They also note the use of a more advanced antenna could make it possible to monitor NFC communications distances of up to a meter. Dhwani features JamSecure security that "uses self-jamming coupled with self-interference cancellation at the receiver" to protect data transmissions. The basis of Dhwani's communication system is an acoustic software-defined radio. The researchers determined that 6 KHz was the lower limit for the system below which typical ambient noise would interfere with transmissions.

NASA Details Software Algorithm That Could Precisely Guide Future Spacecraft Landings
Network World (08/13/13) Michael Cooney

Researchers at the U.S. National Aeronautics and Space Administration (NASA) are developing algorithms designed to land spacecraft safely. One of the key algorithms under development is the Fuel Optimal Large Divert Guidance (G-FOLD) algorithm. G-FOLD is designed to incorporate a spaceship's trajectory, speed, and landing information to guide a ship to a safe landing. The algorithm is being flight-tested in conjunction with Masten Space Systems at the Mojave Air and Space Port. NASA says G-FOLD "autonomously generates fuel optimal landing trajectories in real time and provides a key new technology required for planetary pinpoint landing." The algorithm incorporates the maximum and minimum thrust magnitude, the thrust pointing direction, the glide slope, and the maximum velocity. The landing "maneuver must be computed onboard in real time because the state of the lander cannot be predicted at the start of powered descent phase," NASA notes. Conventional algorithms do not optimize fuel usage and significantly limit how far the landing craft can be diverted during descent, NASA adds.

Blended Learning Improves Test Scores: Study
InformationWeek (08/12/13) Kelly Sheridan

Combining traditional classroom methods with computer-mediated activities increases student test scores, according to a study by the RAND Corporation and the U.S. Department of Education. The study of 18,000 students in 147 high schools and middle schools in seven states showed the effectiveness of blended learning for a high school algebra curriculum. Schools were paired based on achievement profiles and test scores, with half following a traditional classroom-based approach while the other half used software to supplement instruction. The software tracked student problem-solving strategies and offered feedback for improvement. Over the two-year study, students in the software group spent two days per week using the technology, and three days engaging in group work and problem-solving without the software. In the first year, the software group did not show progress, but in the second year test scores rose markedly, with a high school student who would typically score in the 50th percentile scoring in the 58th percentile after a year of using the software. RAND wants to further study the direct impact of the software in the classroom.

Carnegie Mellon Researchers Say Readers’ Identities Can Reveal Much About Content of Articles
Carnegie Mellon News (PA) (08/12/13) Byron Spice

A new study from Carnegie Mellon University (CMU) and the University of Washington indicates that significant information about the content of an article can be gleaned from the attributes of its readers. The researchers analyzed nearly 3 million news articles and public profiles of the Twitter users sharing those articles. Using this information, the researchers created a few thousand badges describing the content of the shared news articles that could also help analyze subsequent articles, even those that had not been shared or read. "Because these badges are based on the readers' self-described interests, rather than only on the words in each article, we found that the badges provide a consistent, reliable means of representing content," says Khalid El-Arini, a recent CMU graduate now research scientist at Facebook. "For instance, while what it means to be 'liberal' changes from month to month, there will always be people who describe themselves as 'liberal' on Twitter, allowing us to produce a direct correspondence between 'liberal' topics from different periods of time." The team extracted the most meaningful, unique words from the articles to develop a "bag of words" for each article that resembles a visual word cloud, assigning more weight to more important words. Descriptive words were similarly taken from Twitter profiles, and the researchers assembled a dictionary associating each badge with its characteristic words.

Researchers Explore Underground Market of Twitter Spam and Abuse
CSO Online (08/14/13) Steve Ragan

A project led by the International Computer Science Institute studied a black market on Twitter in which crooks sell access to accounts that are later used to promulgate spam, malevolent links, inflate follower counts, and launch attacks to suppress political speech. "Revenue generated by miscreants participating in this market varies widely based on business strategy, with spam affiliate programs generating $12 million to $92 million and fake antivirus scammers $5 million to $116 million over the course of their operations," the researchers note. They observe that specialization within this environment is routine, and the manifestation of account merchants represents another specialization "where sellers enable other miscreants to penetrate walled garden services, while at the same time abstracting away the complexities of CAPTCHA solving, acquiring unique emails, and dodging IP blacklisting." Criminal merchants pre-age most of the sold accounts to prevent their instant flagging, evading heuristics that disable new accounts based on weak, early signals of misbehavior. This also helps give the accounts longevity, as older accounts must meet a much higher threshold before being terminated because of potential malicious actions. Among the researchers' recommended remedial actions is increasing the cost of fraudulent account operations, either by focusing on email verification or the use of CAPTCHAs.

Bubbles Are the New Lenses for Nanoscale Light Beams
Penn State Live (08/12/13) Hannah Y. Cheng

Pennsylvania State University researchers have created liquid bubbles to focus nanoscale light beams, which could pave the way for next-generation, high-speed circuits and displays. A team led by professor Tony Jun Huang generated separate models of the light beams and bubble lens to anticipate their behaviors and optimize conditions before integrating the two in the laboratory. The researchers formed the bubble lens by heating water on a gold surface with a low-intensity laser. The bubble's optical behavior maintains consistency if the laser's power and the environmental temperature remain constant. The primary benefit of a bubble lens is the speed and ease of changing its location, shape, and size, which influence the orientation and focus of the light beams. Moving the laser or adjusting its power can change the bubble's deflection of a light beam, either widely dispersing it or focusing it at a specific target. The light beam's refraction also can be altered by changing the liquid. "In addition to its unprecedented reconfigurability and tenability, our bubble lens has at least one more advantage over its solid-state counterparts: its natural smoothness," Huang notes. "The smoother the lens is, the better quality of the light that pass through it."

Your Signature in Lights: Device for Capturing Signatures and Fingerprints Uses Tiny LEDs Created With Piezo-Phototronic Effect
Georgia Tech News (08/11/13) John Toon

Georgia Tech researchers have developed a sensor device that converts mechanical pressure directly into light signals that can be captured and processed optically. The researchers say the technology could provide an artificial sense of touch, offering sensitivity comparable to that of the human skin. The technique also could be used in biological imaging and micro-electromechanical systems by creating a database of signatures and fingerprints. "This is a new principle for imaging force that uses parallel detection and avoids many of the complications of existing pressure sensors," says Georgia Tech professor Zhong Lin Wang. The technology, known as piezo-phototronics, provides a way to capture information about pressure applied at very high resolution. "When you have a zinc-oxide nanowire under strain, you create a piezoelectric charge at both ends, which forms a piezoelectric potential," Wang says. "The presence of the potential distorts the band structure in the wire, causing electrons to remain in the p-n junction longer and enhancing the efficiency of the [light-emitting diode]." He notes the device has the ability to see all of the emitters at the same time, which allows for the quick response time. "When the light emission is created, it can be detected immediately with the optical fiber," Wang says.

The New Age of Algorithms: How It Affects the Way We Live
Christian Science Monitor (08/11/13) Robert A. Lehrman

Big data technology is poised to revolutionize every aspect of human life and culture as people collect and analyze vast volumes of data for behavioral prediction, problem solving, security, and countless other applications. The generation of massive amounts of data is driven by the increasing digitization of everyday activities and people's reliance on electronic devices that leave digital trails. One notable big data project is an effort by the U.S. Library of Congress to cost-effectively archive millions of tweets a day for their historical value. The potential for Google searches to be used to predict local disease outbreaks is the focus of another big data project headed by Johns Hopkins University professor Richard Rothman and colleague Andrea Dugas. The insights gleaned from an analysis of Google search trends could enable medical facilities to better prepare for epidemics. However, big data also has a sinister edge, in the technology's perceived potential to erode privacy, encourage inequality, and promote government surveillance of citizens or others in the name of national security. Such concerns have been thrown into sharp relief amid revelations about the U.S. National Security Agency's PRISM program to secretly monitor communications in the United States and other nations.

Abstract News © Copyright 2013 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