Association for Computing Machinery
Welcome to the August 6, 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).


Animation Research Could Offer Unparalleled Control of Characters Without Skeletons
Georgia Tech News (08/02/12) Joshua Preston

Georgia Tech researchers have developed a way to simulate and control the movement of skeleton-free computer-generated characters. The researchers say their modeling technique could enable amateur animators to control digital creatures with simple point-and-click movements. The computer models, known as soft body ABCs, were created using muscle fibers to control a volume-preserving finite element mesh. The researchers also developed algorithms that allowed high-level goals, such as walking from one point to another or jumping and then regaining balance. "We’ve built a framework where the user or the animator can just click on a point of the soft body and direct the type of movement he or she wants," says Georgia Tech professor Karen Liu. A major hurdle for the researchers was solving how soft body characters would use a balance strategy during movement. "We believe this research contribution is one that can apply broadly to other problems in animation control," Liu says.

In the Olympics of Algorithms, a Russian Keeps Winning Gold
Technology Review (08/06/12) Tom Simonite

Since 2005, Google researcher and former Moscow State University student Petr Mitrichev has led the world in algorithmic programing. Mitrichev won the most recent champion's title in competitive programming, a contest he says offers a rare island of absolutes in a subjective world. "You have a feeling of satisfaction in a contest when you solve a problem," Mitrichev says. "The beautiful aspect is that it's totally automatic and there is no human judgment involved at all." Mitrichev earned his world ranking after joining the premier leagues of competitive programming, a series of weekly and annual contests that attract more than 400,000 programmers from around the world. The problems in these competitions often outline physical situations that must be described mathematically, which requires writing algorithms. Some contests also include a challenge phase, in which the goal is to submit data that trips a competitor's program. At Google, Mitrichev works to improve Google's search algorithm, mostly by making subtle changes, such as finding language tricks to get more meaning from a search term. "When you have an idea, then you implement it as an algorithm, and then you run an experiment," he says.

Lawrence Landweber Helped Build Today’s Internet, Now He’s Advising Its Future
Wired News (08/06/12) Sarah Mitroff

When Lawrence Landweber helped develop an early version of the Internet, known as the Computer Science Network (CSNET), an intentionally open computer network, he predicted that it would one day be used for banking, travel, and commerce. However, Landweber did not foresee a system in which hackers could take down Web sites or extract private information. Landweber also helped build email, directory, and file-sharing software to run on CSNET. "I realized very early that CSNET was going to be international, so starting in 1982 I got together with people all over the world to help them coordinate their networks," Landweber says. Today, Landweber is involved with the Global Environment for Network Innovations (GENI) project, which aims to create faster and more secure networks. He says there is a battle between hackers who steal information, bot computers that create spam, government forces trying to counter attacks, and researchers trying to close security gaps. “The current Internet has serious flaws,” Landweber says. “The Internet was never designed to be secure, and in the formative days, privacy wasn’t that much of an issue.” GENI project researchers are developing software-defined networks in an attempt send data more securely.

Judea Pearl’s Turing Award Lecture at AAAI-12
CCC Blog (08/02/12) Douglas Fisher

University of California, Los Angeles professor Judea Pearl, who received the 2011 ACM A.M. Turing Award, recently gave his Turing Award Lecture at the 26th Association for the Advancement of Artificial Intelligence Conference. Pearl's lecture, entitled "The Mechanization of Causal Inference: A 'Mini Turing Test' and Beyond," highlighted the abilities of humans to do counterfactual reasoning in their exploration of alternatives. Pearl also emphasized the special role of computer science and artificial intelligence in understanding, abstracting, and modeling process, to include human decision processes. He also noted that as a discipline, “we don’t have the luxury of philosophizing” if we are to build intelligent systems. Vanderbilt University professor Douglas Fisher says Pearl's work is a great example of the utility of computational abstractions that bring about profound changes across disciplines. Fisher also notes Pearl's work is a manifestation of how someone with wide-ranging and humanist interests can pursue these with computational abstractions.

Data Storage: Keeping Things in Place
A*STAR Research (08/01/12)

Researchers at A*STAR's Data Storage Institute have developed software that can accurately model the force of air acting on the actuator assembly in a hard disk drive and determine the amplitude of the assembly's vibration. The program is unique because it uses fluid dynamics to describe the interaction between the read/write arm and the air surrounding it. "Variation in the amplitude of vibration can have a critical impact on the performance of hard disk drives, even if the difference is on the sub-nanometer scale," says A*STAR's Ningyu Liu. To create the simulation, the researchers described the space right around the arm as a continuous vibrating surface, instead of dividing it into many cells. The simulations were tested through comparison to measurements of arm vibrations in a home-built simplified hard disk drive. The researchers showed that simulations that fail to include the vibrating boundary can miss the actual amplitude of the vibrating arm by as much as 40 percent.

Mapping the Future of Climate Change in Africa
Texas Advanced Computing Center (08/01/12) Diego Joaquin Cruz Ramirez

Researchers at the Texas Advanced Computing Center (TACC) are working on the Climate Change and African Political Stability (CCAPS) program, which features an online mapping tool that analyzes how climate and other factors interact to threaten the security of African communities. "The first goal was to look at whether we could more effectively identify what were the causes and locations of vulnerability in Africa, not just climate, but other kinds of vulnerability," says University of Texas at Austin professor Francis J. Gavin. CCAPS consists of nine research teams focusing on different aspects of climate change, their relationship to different types of conflict, the government structures that exist to mitigate them, and the effectiveness of international aid in intervening. The researchers, led by University of Texas at Austin professor Joshua Busby, examined four different sources and then combined them to form a composite map. Busby and his team relied on three regional climate model simulations that took about two months to complete on TACC's Ranger supercomputer. "The mapping tool is a key part of our effort to produce new research that could support policy making and the work of practitioners and governments in Africa," says CCAPS's Ashley Moran.

Learning Machines Scour Twitter in Service of Bullying Research
University of Wisconsin-Madison (08/01/12) Chris Barncard

University of Wisconsin-Madison researchers are developing a computational method for searching through social media posts to find mentions of bullying events. "What we found, very importantly, was that quite often the victim and the bully and even bystanders talk about a real-world bullying incident on social media," says Madison professor Jerry Zhu. The computer analyzed about 250 million publicly visible messages posted on Twitter and found more than 15,000 bullying-related tweets per day. "We taught [the machine-learning algorithm] ways to identify bullies, victims, accusers, and defenders," says Madison professor Amy Bellmore. The technique also could be used to identify children in need of intervention. "The idea is that if someone is powerfully affected by the event, if they are feeling extreme anger or sadness, that's when they could be a danger to themselves or others," Zhu says. Using the data to show victims that they are not alone also could help children deal with their feelings, according to the researchers. "When they're exposed to the idea that other people are bullied, actually it has some benefit," Bellmore says.

Massive Data for Miniscule Communities
MSU News (08/01/12) Layne Cameron

Michigan State University (MSU) researchers have developed a computational technique that relieves logjams that commonly occur in big data sets. The researchers note that microbial communities' genomic data is easy to collect, but the data sets are so large that they can overwhelm conventional computers. "To thoroughly examine a gram of soil, we need to generate about 50 terabases of genomic sequence--about 1,000 times more data than generated for the initial human genome project," says MSU professor C. Titus Brown. He notes the strategy is unique in that it was created using small computers rather than supercomputers, which is the usual approach for bioinformatics research. The method utilizes a filter that folds the data set up using a special data structure, which enables computers to analyze small portions of the data at a time. The technique creates a 40-fold decrease in memory requirements, enabling scientists to sift through large volumes of data without the use of a supercomputer. The researchers made the technique's source code publicly available to encourage others to extend it.

Scientists Mimic Guitar Hero to Create Subliminal Passwords for Coercion-Proof Security (08/01/12) Gareth Morgan

Stanford University researchers have developed a security system that can defeat the most aggressive attackers by using subliminal passwords. To create the subliminal password, the researchers developed a computer training game, similar to the popular video game Guitar Hero, in which players try to time circles falling into columns and press keys that correspond to those columns at just the right time. The subliminal password is created from a sequence of 30 characters using that set of keys, although users are never told what the sequence is. The users are presented with several training sessions, which include the 30-character sequence, as well as other random sequences. To authenticate the user, the subject is presented with some circles that fall in the patterns used during the training session, and others that were not used. The researchers showed that users consistently scored better on the sequences they had been trained to implicitly learn than the random ones. "We hope to further analyze the rate at which implicitly learned passwords are forgotten, and the required frequency of refresher sessions," the researchers say.

National Lab Replaces Supercomputer With Newer, Faster Model
InformationWeek (07/31/12) Patience Wait

Argonne National Laboratory recently started accepting applications from scientists that want to use its new Mira supercomputer, which is ranked the third fastest in the world and has 768,000 core processors and operates at more than eight petaflops. Mira's initial applications include studying the quantum mechanics of new materials, measuring the role and impact of clouds on climate, and modeling earthquakes. Those and 13 other projects are part of Argonne's Early Science Program and are intended advance science, as well as evaluate Mira's performance, according to Argonne's Mike Papka. "A new architecture with a new system software stack, and at a scale that is larger than anyone else has run previously, results in a system that will have issues never seen before," Papka says. "These issues need to be exposed and addressed before we go into production, and it often requires real users running real code on the system." About 60 percent of Mira's processing cycles will be devoted to projects selected for the U.S. Department of Energy's Innovative and Novel Computational Impact on Theory and Experiment program, and 30 percent will go to projects accepted into the Advanced Science Computing Research Leadership Computing Challenge.

Stanford's Hottest Major: Computer Science
Palo Alto Online (CA) (07/31/12)

More than 220 Stanford University students, in a class of about 1,700, chose computer science as their major last year, a 25 percent increase from the previous record in 2000-2001, according to a Stanford School of Engineering report. The increase in computer science students at Stanford follows a major restructuring of the program over the past several years. The new program includes six courses that provide a foundation, which is built on a series of tracks that students can choose from in order to focus on their own personal interests. The tracks include artificial intelligence, systems, theory, graphics, and human-computer interaction. Students also can include courses from other departments, such as biology, studio art, and psychology, as part of their computer science program. "Virtually every field is touched by computer science in some way," notes Stanford computer science department associate chair Mehran Sahami. "In medicine and biology computational methods are used to analyze DNA, predict treatment outcomes, and model drugs at a molecular level."

Rutgers Engineers Design Cell Phone App to Reduce Distracted Driving
Rutgers University (07/31/12) Carl Blesch

Researchers at Rutgers University and the Stevens Institute of Technology have developed a smartphone application that can identify where a cell phone user is sitting in a car and automatically adjust the phone's settings to try to keep the driver's attention on the road. "We're making it easier for people who want to drive less distracted," says Rutgers professor Marco Gruteser. For example, the app can silently forward incoming calls and texts to message boxes for later retrieval. The app also could automatically respond to a caller or texter, saying the owner is currently driving and will reply later. For outgoing communication, the app could disable texting and make placing certain frequent calls less difficult by displaying them as large on-screen buttons. The app works with the car's sound system to distinguish between the driver and passenger, and requires a stereo sound system with Bluetooth connectivity. Beeps come out of left and right speakers at different intervals, and the phone uses its microphone to listen for the beeps it just sent. If beeps from the left speaker arrive fractions of a second faster than beeps from the right speaker, it means the phone is likely in the driver's hands.

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