Welcome to the June 15, 2015 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
Chinese Hackers Circumvent Popular Web Privacy Tools
The New York Times (06/12/15) Nicole Perlroth
New research indicates Chinese hackers have found a way to circumvent virtual private networks (VPNs) and the Tor anonymous browsing network to target users who visited websites frequented by Chinese journalists and the Uighur ethnic minority. VPNs and Tor are two of the most trusted tools used to circumvent China's Great Firewall. Although it is not clear who the hackers are, the nature of the targets and the sophistication of the attacks suggest they are related to the Chinese government. According to AlienVault security researcher Jaime Blasco, the hackers were able to exploit a known security vulnerability called JSONP to circumvent VPNs and Tor and launch watering-hole attacks on the target websites. If users were logged into any one of China' top 15 Web portals, such as Baidu or Taobao, at the time, the hackers were able to capture personal information, including names, physical and email addresses, phone numbers, birth dates, and the users' browsing activity. The JSONP vulnerability exploited by the hackers has been known since 2013, and Blasco says by failing to patch it, the major Web portals effectively neutered whatever privacy protections were available to Chinese users. "There's a growing sense within China that widely used VPN services that were once considered untouchable are now being touched," says Berkman Center for Internet and Society at Harvard fellow Nathan Freitas.
Bridging the Rift Between Classroom and Online Learning
Penn State News (06/15/15) Katie Jacobs
Pennsylvania State University (PSU) researchers are experimenting with Oculus Rift, an immersive virtual reality (IVR) system, to determine how it could enhance student learning. The researchers say IVR technology eventually could give students taking online courses a way to become more engaged in their coursework. PSU professor Conrad Tucker says although online courses provide important opportunities in higher education, they also are limiting because there is little immersive or tactile interaction, which makes it hard for students to engage with the material; he believes IVR systems could be a solution to that problem. The PSU researchers, working with the Oculus Rift IVR headset, designed simulations using the Unity3D program, as well as a haptic glove associated with the IVR system. The study found the PSU system significantly improves a student's performance in completing a task when compared to doing the same activity in a non-immersive computer program. Although this type of system has many benefits, "one of the major ones is that when compared to the non-immersive system, IVR systems give you a much more natural experience," Tucker says.
Deep-Learning Machine Beats Humans in IQ Test
Technology Review (06/12/15)
University of Science and Technology of China (USTC) researchers say they have built a deep-learning machine that outperforms the average human's ability to answer verbal reasoning questions. Traditionally, computer science researchers have used data-mining techniques to analyze huge volumes of texts to find the links between words they contain; this approach assumes each word has a single meaning represented by a single vector. The USTC researchers solved this problem by taking each word and seeking other words that often appear nearby in a large corpus of text, and then using an algorithm to see how these words are clustered. They then looked up the different meanings of a word in a dictionary and matched the clusters to each meaning, with the overall result being a way of recognizing the different senses that some words can have. The researchers also developed a method to make it easier for a computer to answer verbal reasoning questions by identifying the category of each question so the computer then knows which answering strategy it should employ. The researchers then developed an algorithm for solving each one using the standard vector methods.
Interpreting Ultrasound Images Using an App
Gemini (06/15/15) Havard Egge
Researchers at SINTEF, the largest independent research organization in Scandinavia, have developed a game-like app that trains health-care professionals to interpret ultrasound images. Ultrasounds can now image nearly all parts of the body, and those images can be used to study various organs and glands, as well as for assistance and guidance in different types of surgical operations. "One major advantage of ultrasound is that it makes a number of examinations and operations simpler and less time-consuming, as well as being easier on the patient," says SINTEF researcher Kaj Johansen. The game has three levels of difficulty, so new tasks appear as users become better at the game. On the first level, users must identify the artery shown in red. On the most difficult level, the user is given the same task, but no help is given and the task is timed. The app also trains users in identifying a nerve as well as the surrounding landmarks in ultrasound images. One of the tasks in the game is to place a needle close to the nerve and inject anesthetics around it prior to an operation on the leg. Although only a limited amount of illustrative material has been incorporated in the beta version of the app, "Our aim is to add more data to the app that will cover more of the anatomical variability seen in patients," says SINTEF researcher Frank Lindseth.
Robotics Competition Generated Groundbreaking Research
MIT News (06/11/15) Larry Hardesty
The Massachusetts Institute of Technology (MIT) team finished sixth out of 25 contestants in the final round of competition in the U.S. Defense Advanced Research Projects Agency's contest to design control systems for a humanoid robot that could perform several tasks related to a hypothetical disaster. When a bipedal robot takes a step, its foot strikes the ground at several points, which experience different forces over time. A function that factors in all of those forces would be difficult to optimize, but it becomes more feasible if the forces are treated as acting on each foot at a single point. The MIT researchers solved that problem by developing a planner that factors in contacts between the robot's arms, the objects the robot is manipulating, and the surrounding environment. In addition, the planner accounts for the forces exerted by those contacts in six dimensions rather than three. The MIT planner then combines all of these factors together into one optimization problem, choosing just those points of contact that minimize displacement of the robot's center of gravity, while still accommodating environmental constraints. The MIT robot also is equipped with sensors that identify objects in the area. The algorithm automatically determines the extent of safe areas by locating the first significant changes in altitude.
Cyber Citizen Tool Shows Which Countries' Laws Cover Our Surfing
New Scientist (06/11/15) Aviva Rutkin
London artist James Bridle wants to bring attention to the way the Internet is making people more international, or even super-national, and he has created a new tool that can determine how much time users spend on different countries' websites and then compute their algorithmic citizenship. Called Citizen Ex, the tool uses Internet Protocol addresses to map where users are and where each website they visit might be based. Citizen Ex forms a picture of the geographical spread of user browsing over time based on the hosting country of each website visited and the amount of time spent on that site. Bridle says a person's algorithmic citizenship is important because it can determine the laws people are subject to and how authorities view them. "It's a form of citizenship that's actually being enacted right now," he says. "I wanted to build something that made it visible." Bridle notes intelligence agencies may consider spying on anyone with sufficiently foreign browsing habits. "The promise of the Internet as an open, free, and borderless place relies on you being able to take advantage of certain tools that enhance your privacy," he says.
Facebook Tool Could Help Devs Pinpoint Elusive Memory Leaks
IDG News Service (06/11/15) Joab Jackson
Facebook has released its Infer static analysis debugger as an open source tool. The current version can examine code written in C, Objective-C, and Java, but Facebook hopes others will add support for more languages. Organizations and individuals can use Infer to build mobile and desktop apps, according to Facebook developer Peter O'Hearn, one of the creators of the tool. Static analyzers test programs by mathematically examining code for all possible ways it can run, flagging incorrect or unfinished bits. The software can separate large applications into a set of smaller parts so they can be examined more easily. Moreover, Infer can remember parts it has analyzed already and will not recheck them unless they have been modified. O'Hearn says the tool also is well adapted to sniffing out memory leaks and null pointer access. Facebook has used Infer to inspect Facebook Messenger, Instagram, and its apps for the Android and iOS platforms.
Stanford Engineers Team Up With U.S. Army to Set Computational Record
Stanford Report (06/10/15) James Urton
Engineers at Stanford University collaborated with the U.S. Army Research Laboratory (ARL) to set a computational record. Stanford professor Charbel Farhat and his research team at the Army High Performance Computing Research Center (AHPCRC) used a massively parallel computer to demonstrate the power of algorithms that instruct processors to solve challenging problems as a group. A total of 22,000 processors were used to solve more than 10 billion mathematical equations in a little over three minutes. Farhat's team collaborated with the ARL Defense Computing Resource Center to craft algorithms that divide calculations among thousands of computers. Team members worked continuously for three weeks to prepare their software for the test on the Excalibur Cray XC40 computer. On the day of the test, they had access to a significant portion of the facility's 101,184 processors to divide their equations into components, share information, and solve the problem. The entire Excalibur will likely not be available for a similar exercise, but Farhat says "when we improve our algorithms on the highest-end computers, we get to benefit from these improvements on our more pedestrian computers."
Robotic Harvesting of Broccoli Could Be Coming to a Field Near You
University of Lincoln (06/10/15) Daniels, Marie
A project involving three-dimensional (3D) camera technology at the University of Lincoln could result in a fully automatic robotic harvesting system for broccoli. The project is being funded by Agri-Tech Catalyst, the Biotechnology and Biological Sciences Research Council, and Innovate U.K., and will test whether 3D camera technology can be used to identify and select when broccoli is ready for harvesting. The research team consists of Tom Duckett and Grzegorz Cielniak from Lincoln's School of Computer Science and Simon Pearson from the university's National Center for Food Manufacturing (NCFM). Industry partner R. Fountain & Son will be responsible for creating the broccoli-cutting device. Another project benefiting from the university's expertise is the early detection and biocontrol of prevalent diseases of mushrooms and potatoes. This project addresses how diagnostic tools can help identify, prevent, and manage disease and find alternatives to chemical pesticides. "At the heart of the project is a drive to develop robust solutions for bio-monitoring and bio-control, leading to scientific advancement and the marketing of products which will ultimately have significant economic and societal benefit for the U.K. and beyond," says the NCFM's Bukola Daramola.
Newly Discovered Property Could Help Beat the Heat Problem in Computer Chips
Stanford Linear Accelerator Center (06/10/15)
Researchers at the U.S. Department of Energy's Stanford Linear Accelerator Center (SLAC) National Accelerator Laboratory observed an exotic property that could alter the electronic structure of a material to reduce heat buildup and improve performance in computer components. The researchers studied a form of iridium oxide and found it has a long-theorized property called 3D negative electronic compressibility. This means the material's electronic structure, instead of its physical structure, substantially warps as electrons are added. The researchers found that a gap between different groupings of energy bands in the sample material shrank as electrons were added, reducing the material's stored energy level. In theory, using this type of material in transistors could substantially enhance their efficiency and reduce heat buildup, according to Boston College researcher Junfeng He. The researchers currently are working on the first demonstration of the material's potential application to transistors. They precisely measured the electronic structure of the material using an advanced x-ray technique at SLAC's Stanford Synchrotron Radiation Lightsource. "This work informs us of the importance to continually look for other new materials with novel physical properties for use in transistors and for other applications," says University of California, Santa Barbara professor Stephen Wilson.
Microsoft Algorithm Improves Directions in Large Networks for Bing Maps
Microsoft researchers say a project for Bing Maps led to the development of the first routing engine to satisfy many algorithmic requirements for multi-stage trips. The researchers say Customizable Route Planning more accurately estimates the time needed for turns, U-turns, road closures, and traffic jams. The research leveraged a classic algorithm by 1972 ACM A.M. Turing Award recipient Edsger Dijkstra that was thought to perform too slowly to calculate online map routes. The authors found no previous technique met all the requirements for a real-world routing engine, such as the ability to incorporate comprehensive road network data; factoring in things like shortest distance, walking, biking, avoiding U-turns, height and weight restrictions, and other problems that pop up; and making rapid calculations to match real-time traffic information with information about the roads. The engine also must support not only the calculation of point-to-point shortest paths, but also suggest several alternate routes. The researchers say they turned to separator-based methods instead of those that exploit the hierarchical structure of roads, and combined them with careful engineering to significantly improve performance and enable interactive applications. The explicit separation of metric customization from metric-independent preprocessing also enabled Bing Maps to answer arbitrary questions about a trip in milliseconds.
Listen to Me: Machines Learn to Understand How We Speak
The Conversation (06/11/15) Michael Cowling
During its recent World Wide Developer Conference, Apple announced additional features it was adding to the voice recognition capabilities of its Siri personal assistant app as part of its most recent iOS 9 update. One of Siri's new abilities is a limited capacity to understand context, for example telling Siri to "remind me of this" while viewing a Facebook invite. Voice recognition software has come far in recent years, especially with the introduction of natural-language processing and artificial neural networks that can be trained to recognize language. Google most recently reported error rates of less than 8 percent. However, major challenges still exist, writes Central Queensland University senior lecturer Michael Cowling. He notes pronunciation is a significant problem, especially in a language such as English where pronunciation does not always align perfectly with spelling and can vary depending on a person's accent or dialect. Software also struggles to pick up on contextual clues people easily recognize. Still, Cowling says progress is being made every day. Both Microsoft and Google recently revealed impressive advancements in automatic translation. Google, for example, unveiled technology that combines image or voice recognition, natural-language processing, and a smartphone camera to automatically translate signs or short conversations.
Abstract News © Copyright 2015 INFORMATION, INC.
To submit feedback about ACM TechNews, contact: firstname.lastname@example.org
Current ACM Members: Unsubscribe/Change your email subscription by logging in at myACM.