Association for Computing Machinery
Welcome to the November 17, 2014 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).


U.S. Sets Sights on 300 Petaflop Supercomputer
Computerworld (11/14/14) Patrick Thibodeau

The U.S. Department of Energy on Friday announced plans to spend $325 million on a pair of supercomputers that could help the United States retake the lead in supercomputing power. China currently is home to the world's fastest supercomputer, the 34-petaflop Tianhe-2. The new proposed supercomputers, which would be housed at Oak Ridge and Lawrence Livermore National Laboratories, would be 150-petaflop machines that potentially could be boosted to 300 petaflops with additional investments. The new supercomputers, to be named Summit and Sierra, respectively, will be built using IBM Power central-processing units and NVIDIA's still-in-development Volta graphics-processing unit. The machines will represent not just a major jump in compute power, but also energy efficiency: the Oak Ridge system is expected to consume the same amount of power as the lab's existing Titan supercomputer, but will run five times faster. The new computers were praised by U.S. Reps. Chuck Fleischmann (R-TN), Dan Lipinski (D-IL), and Bill Foster (D-IL) as helping the U.S. reclaim its lead in the field of high-performance computing. Foster says he has high hopes for further investments in supercomputing research. "Supercomputing is one of those things that we can step up and lead the world again," Fleischmann says.

Getting Serious About the Design of Social Computing Systems
CCC Blog (11/13/14) David W. McDonald

The difficulties of designing social computing systems come from the complexity of the software and hardware configurations, as well as the fact that some participants in a social computing system will not behave with positive goals or intent, writes David W. McDonald, chair of the University of Washington's Human Centered Design and Engineering Department. He says as participation rates in social computing system grow, system designers must work to prevent or mitigate the potential risks these systems pose for the participants and to society. Another challenge associated with social computing is exploring the design risks. McDonald says social computing systems represent a complex combination of human behaviors, human interactions, algorithms, and collections of hardware and devices resulting in a design space that is necessarily uneven and irregular, making some design choices difficult to fully comprehend. Often, because designers focus on positive outcomes with positive assumptions about user behaviors, they only explore a small portion of the total design space, leaving potential risks unclear. Finally, McDonald says there are great potential societal benefits to social computing, including more effective crisis response, safer neighborhoods, more satisfying work and family relationships, and responsive education and workforce training.

Magic Tricks Created Using Artificial Intelligence for the First Time
Queen Mary, University of London (11/17/14) Will Hoyles

Queen Mary, University of London artificial intelligence (AI) researchers have taught a computer to create magic tricks. "While a member of the audience might have seen a variation on this trick before, the AI can now use psychological and mathematical principles to create lots of different versions and keep audiences guessing," says Queen Mary researcher Howard Williams. For example, the magic jigsaw puzzle involves assembling the pieces to show a series of shapes, then taking it apart and reassembling it so the shapes have disappeared using a geometric principle. This type of trick involves several simultaneous factors, such as the size of the puzzle, the number of pieces involved, the number of shapes that appear and disappear, and the ways the puzzle can be arranged. For the mind-reading card trick, a computer was used to arrange the decks in such a way that a specific card could be identified with the least amount of information possible. "Using AI to create magic tricks is a great way to demonstrate the possibilities of computer intelligence and it also forms a part of our research into the psychology of being a spectator," says Queen Mary professor Peter McOwan.

Elementary Schools Start Teaching Data Literacy
The Washington Post (11/16/14) Mohana Ravindranath

Elementary school teacher Lisa Parisi is one of several teachers around the country who are helping to address the U.S.'s shortage of qualified data scientists by teaching data literacy to students at an early age. Her fifth-grade students at the Denton Avenue School in New Hyde Park, NY, make use of data sets in their social studies courses. Meanwhile, students at the Maury Elementary School in Washington, DC, gather their own data, such as daily temperatures and the movement of the sun, to enrich their science projects. Experts say these basic lessons could help students develop the data literacy skills needed to meet the U.S.'s shortage. Consulting firm McKinsey estimates the U.S. is facing a shortage of 140,000 to 190,000 professionals with analytical expertise and 1.5 million analysts and managers that can handle big data analysis. "It makes sense for us to be thinking about education, starting in early childhood, about concepts such as the difference between correlation and causation, what it means to have a bias as you think about data, conditional probability," says McKinsey's Michael Chui. He says these lessons will help students to learn the limitations of data as well as its powerful potential.
View Full Article - May Require Free Registration | Return to Headlines | Share Facebook  LinkedIn  Twitter 

Recommendation Theory
MIT News (11/14/14) Larry Hardesty

Massachusetts Institute of Technology (MIT) researchers have applied a standard algorithm, known as "collaborative filtering," to the recommendation engines that are used on websites such as Amazon and Netflix. To generate specific recommendations, the algorithm first assigns other users similarity scores based on the degree to which their ratings overlap with another user. The algorithm then would aggregate the ratings a movie received from other users, weighted to similarity scores. The researchers simplified the analysis by assuming the ratings system had two values, either positive or negative, so the preferences of every user could be described by a string of ones and zeros. The MIT model assumes that large groups of such strings can be clustered together, and those clusters can be described probabilistically. The researchers showed that as long as the number of clusters required to describe the variation in a population is low, collaborative filtering yields nearly optimal predictions. The researchers studied 10 million users of a movie-streaming site and identified 200 who had rated the same 500 movies. The researchers found that just five probabilistic models were enough to account for most of the variation in the population.

Self-Repairing Software Tackles Bugs
University of Utah News (11/13/14) Vincent Horiuchi

University of Utah researchers have developed the Advanced Adaptive Applications (A3) software suite to detect and eliminate malware. They say A3 instantly repairs damage and prevents malware from infecting the computer again. A3 works with a virtual machine that emulates the operations of a computer without dedicated hardware. The software is designed to monitor a virtual machine's operating system and applications, notes University of Utah professor Eric Eide. A3 is intended to protect servers or similar business-grade computers that run on the Linux operating system, along with military applications. There are no plans to adapt A3 for home computers or laptops, but Eide says this could be a possibility in the future. A3 can detect new, unknown viruses or malware automatically by sensing that something is occurring in the computer's operation that is not correct. A3 stops the virus, determines a repair for the damaged software code, and learns to permanently block that malware from entering the machine. To test A3's effectiveness, a team from the university and Raytheon BBN used it against the Shellshock malware. A3 was able to discover Shellshock on a Web server and repair the damage in four minutes, while six other pieces of malware also were successfully handled by A3, Eide notes.

IBM Shares Plans for Supercomputing Future
IDG News Service (11/13/14) Agam Shah

IBM is developing a new supercomputing architecture equipped with more co-processors and accelerators to increase computing speed and power efficiency. The aim is to boost data processing at the storage, memory, and input/output levels, according to IBM's Dave Turek. He says the new architecture will help break down parallel computational tasks into small chunks, reducing the compute cycles required to solve problems. "When we are working with petabytes and exabytes of data, moving this amount of data is extremely inefficient and time-consuming, so we have to move processing to the data," Turek says. "We do this by providing compute capability throughout the system hierarchy." He notes the size of the data sets can be reduced by decomposing information in storage, which can then be moved to memory. "We see a hierarchy of storage and memory including nonvolatile RAM, which means much lower latency, higher bandwidths, without the requirement to move the data all the way back to central storage," Turek says. He notes IBM is now looking at optimizing entire supercomputing workloads, which involve modeling, simulation, visualization, and complex analytics, on massive data sets. "Our own research shows that many classic [high-performance computing] applications are only moderately related to the measure of LINPACK," a benchmark measurement based on floating point operations, Turek says.

Study Finds Search Engine Poisoning Persistent, Hard to Solve
eWeek (11/13/14) Robert Lemos

Efforts to prevent the manipulation of search results are having little impact on fraud, according to a study by researchers at Carnegie Mellon (CMU) and Southern Methodist universities. The researchers say Google's countermeasures during their four-year study were not overly successful, and they believe search-redirection attackers have improved their techniques. The study focused on fraud-prone search subjects, such as pharmaceuticals, antivirus, pirated software, and online gambling. Nearly 60 percent of searches for such subjects redirected users to fraudulent sites in 2012. Although the average time to clean the infected systems behind the attacks eventually shrank (to about 15 days), fraudsters compensated by penetrating more systems, says CMU professor Nicolas Christin. Starting in mid-2012, Google released browsers that defaulted to encrypted search requests using secure HTTP, which likely had more impact. "The move to encrypted search meant that certain parameters were not available to the attackers," Christin notes. However, he says cybercriminals may have improved their methods. The researchers believe industry and government should focus their efforts on specific hosting providers and networks that are home to traffic brokers. They say by integrating the elimination of black-hat search results from the search database and a coordinated action against the traffic brokers, defenders could have a more long-term effect.

Researchers Test First 'Smart Spaces' Using Light to Send Data
Dartmouth College (11/12/14)

A key problem in the area of visible light communication (VLC) is that data transmission is blocked if light is obstructed by people's movements, shadows, and other factors. To tackle the problem, Dartmouth College professor Xia Zhou and colleagues are experimenting with "smart spaces" featuring algorithms, ceiling-mounted light-emitting diodes (LEDs), and light sensors in floors and in smart devices. Such smart spaces monitor users' gestures and can separate shadows from light. The researchers' integrated visible light communication project (iVLC) system embeds identifiers into light signals from LEDs, enabling light sensors on the floor to measure individual light sources. Shadows are used to identify user gestures and to track movements to create an uninterrupted flow of data in indoor places. Preliminary results from simulation tests are promising, and the researchers currently are working on a proof-of-concept iVLC testbed using LEDs and light sensors. VLC technologies work by changing the intensity and frequency of LEDs to manipulate optical signals and rapidly send large amounts of data economically and safely. Recent VLC research has focused on using clusters of LEDs in ceilings and receivers in computers.

Fighting Crime Through Crowdsourcing
University of Miami (11/11/14)

University of Miami (UM) researchers are developing a computing model that uses crowdsourcing to combine and optimize human efforts and machine-computing elements. The model uses social networks to perform complex tasks of face recognition and could be a formal part of the criminal investigation process, according to UM vice provost Brian Blake. "The breadth of the Internet and popularity of smartphones have facilitated the onset of online crowdsourcing platforms," Blake says. "Our project attempts to leverage the power of the crowd to solve complex problems, on demand." He notes by combining the efforts of both machine-computing and human-computing elements in performing the same task, there was an overall average certainty of 69.13 percent. Blake says the system is unique because of its elasticity, enabling it to adapt to changes in the workload of a task. He notes the model also allows the machine- and human-computing resources to change, when and where needed and without disrupting the operations. "Elastically, we would like to decide who is best for a specific task, or what concentration of people or machines could be mixed for a specific task," Blake says.

Cybersecurity Skills Need Boost in Computer Science Degrees
Times Higher Education (11/12/14) Chris Parr

U.K. computer science programs are not doing enough to develop students' skills in cybersecurity, according to a new paper from the Council of Professors and Heads of Computing and the International Information Systems Security Certification Consortium (ISC)2. As a result, the United Kingdom faces a shortage in cybersecurity skills. Industry increasingly has called for making cybersecurity a core discipline of computing and information technology, indicating it should be "a key element of the computing and computer science curriculum, particularly at the undergraduate level." The paper notes most institutions offer computer science courses with one module or unit--approximately 5 percent of total credits--dedicated to cybersecurity in a three-year degree. Cybersecurity topics could be included in many degrees. Moreover, the paper suggests industry could accept more trainees and placement students. Industry also could send speakers to share their "war stories" and help get more students excited about cybersecurity.

Scientists Build a Better Eye on Our World
UIC News Center (11/10/14) Jeanne Galatzer-Levy

The University of Illinois at Chicago's Electronic Visualization Laboratory (EVL) has received a $3-million U.S. National Science Foundation grant to develop a video camera that can capture images in 360 degrees and three dimensions (3D). EVL co-founder Daniel Sandin is working on the Sensor Environment Imaging (SENSEI) Instrument, which will capture the whole sphere of an object's environment, in stereo, with calibrated depth, while providing information on how far away other objects are and allowing estimates of their size and mass. Sandin says the video image exhibits true colors and brightness. To create a composite view of multiple images, "you have to stitch together pictures that overlap and are taken at different angles," says EVL director Maxine Brown. She notes SENSEI is video, "which means not one picture, but 30 per second, in stereo, and from multiple cameras, not only set at different angles but sitting at different points." A successful prototype will require the developers create a configurable, portable, sensor-based camera system capable of handling a large amount of data, displays for viewing the 3D stereoscopic images, and network systems that enable collaboration. SENSEI also will require hardware scaffolding for its sensor arrays, data acquisition, computing platforms, telemetry, communications, and software. Potential users include oceanographers, computer scientists, astronomers, archaeologists, and public health experts.

Abstract News © Copyright 2014 INFORMATION, INC.
Powered by Information, Inc.

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