Association for Computing Machinery
Welcome to the October 16, 2009 edition of ACM TechNews, providing timely information for IT professionals three times a week.

HEADLINES AT A GLANCE


What Kind of Cloud Computing Project Would You Build With $32M?
Network World (10/14/09) Cooney, Michael

The U.S. Department of Energy (DOE) plans to build a large cloud computing test bed in an effort to determine whether cloud computing can help meet scientists' demand for computing resources. Approximately $32 million will be spent on the Magellan project, which will combine the commercial cloud offerings of Amazon, Microsoft, and Google. The project also will link the 100Gbps Energy Sciences Network (ESnet) to the Argonne National and Lawrence Berkeley National laboratories to rapidly transfer data between geographically dispersed clouds. ESnet will enable DOE scientists to access the computing resources regardless of their location. As DOE scientists use the Magellan system for their computations, performance-monitoring software will be used to analyze the kinds of science applications being run on the system and how well they perform on a cloud. The researchers say the project will help provide a better understanding of cloud computing's potential as a cost-effective and energy-efficient tool for scientific discovery. "We know that the model works well for business applications, and we are working to make it equally effective for science," says Argonne's Pete Beckman.


Carnegie Mellon Researchers Save Electricity With Low-Power Processors and Flash Memory
Carnegie Mellon News (10/14/09) Spice, Byron

During ACM's recent Symposium on Operating Systems Principles, researchers from Carnegie Mellon University (CMU) and Intel Labs Pittsburgh (ILP) won the best paper award for their article on Fast Array of Wimpy Nodes (FAWN), a server architecture that can handle data-heavy applications with greater speed and efficiency than current systems. To create FAWN, CMU computer scientist David Anderson and colleagues collaborated with ILP to integrate netbook processors with flash memory. Flash memory works more quickly than hard disks, costs less than DRAM chips, and is the most energy efficient option available. To test the server architecture, the researchers constructed a FAWN computing cluster out of 21 nodes, each equipped with a low-cost, off-the-shelf processor and a four-gigabyte flash card. They discovered that the FAWN cluster could manage 10 to 100 times more requests than a disk-based one while using the same amount of energy. Researchers are now constructing a FAWN cluster that uses Intel's Atom processor. "FAWN systems can't replace all of the servers in a data center, but they work really well for key-value storage systems, which need to access relatively small bits of information quickly," Anderson says. He says that in the future the researchers hope to use FAWN for data analysis.


Merging Video With Maps
Technology Review (10/14/09) Kremen, Rachel

Microsoft and researchers from the University of Konstanz in Germany are collaborating to create Videomap, navigation software that incorporates videos of driving routes. The program gives drivers visual cues by highlighting landmarks and emphasizing one side of the road before a turn. Videomap uses algorithms for "turn anticipation"--essentially, the video slows before a turn and points out key images where the turn must be made. The program points out landmarks in the same way. "As we pass a landmark, the field of view will expand to encompass that landmark and create a landmark thumbnail," says Microsoft's Billy Chen. The image is held for a few seconds so that the driver can commit it to memory. Video speed varies depending on whether the driver wants to note landmarks or get an idea of the length of the trip. To test the system, 20 volunteers read normal driving instructions for five minutes. Then they were shown a simulation of the route and were asked several times to state where the car would turn next. The second time participants used Videomap instructions. With normal directions, the drivers were correct 60 percent of the time; with Videomap, the number rose to 80 percent. Chen calls the study "pretty conclusive," and points out that drivers relied less on text instructions after using Videomap and most of them preferred the software. Chen plans to test participants a second time using a new video simulation to see how the program holds up in different environments. He also wants to develop the program so that users will look only at the video when it covers a landmark, rather than looking equally at both the moving map and video. University of Zurich researcher Arzu Coltekin says that Videomap could potentially be useful for bikers and pedestrians as well.


Seeking Privacy in the Clouds
Duke University News & Communications (10/13/09) Basgall, Monte

Duke University professor Landon Cox recently received a three-year, $498,000 U.S. National Science Foundation grant to research alternatives for providing social networking services that do not concentrate all user information in a single place. Cox believes creating a peer-to-peer architecture to spread the information out would make individual data harder to steal or exploit. "The basic idea is that users would control and store their own information and then share it directly with their friends instead of it being mediated through a site like Facebook," he says. In a report for ACM's Workshop for Online Social Networks, Cox proposed three possible options. In each option, users would load personal information into a virtual individual server (VIS), which could be hosted on the user's computer or be distributed within redundant clouds of servers. One of the options is called hybrid decentralization, which would keep VISs on desktops when possible, but switch to the cloud distribution option when individual computers go offline. "Users can try to put their information in clouds of servers, which are going to be highly available but expensive," Cox says. "Or they could try to store it on their own machines, which would be cheap but subject to service interruptions."


NCWIT Award for Aspirations in Computing
National Center for Women & Information Technology (10/16/09)

The National Center for Women & Information Technology (NCWIT) says the 2010 NCWIT Award for Aspirations in Computing will include affiliate award programs in Texas, Illinois, and Florida. Young women in high school who apply from these states will be entered into both the local and national competitions. A grant from the Motorola Foundation has made the affiliate award programs possible. NCWIT created the Award for Aspirations in Computing to recognize the computing-related achievements of young women, generating the kind of visibility and support that will encourage female students to pursue their computing-related interests. Award winners are selected for their outstanding aptitude and interest in information technology and computing, solid leadership ability, good academic history, and plans for post-secondary education. Winners and their school receive an engraved award. Winners also receive a trip to attend the Bank of America Technology Showcase and Awards Ceremony, which takes place March 27, 2010, in Charlotte, N.C.; $500 in cash; and a laptop computer.


In Search of Machines That Play at Being Human
Carlos III University of Madrid (Spain) (10/14/09)

This year, 15 teams from Brazil, Canada, the United States, Italy, Japan, the United Kingdom, and Spain, among others, participated in the BotPrize contest, which applies the Turing Test to video games. In the competition, a judge started a game against two players, one human and one artificial. After 15 minutes of play, the judge had to identify which player was a human and which was a program. None of the computer programs, or bots, entered in this year's contest was able to deceive 80 percent of the judges. "In our case, we didn't have enough time to program a good bot, since I am still in the process of migrating the control architecture that I use in real robots to the bots in the Unreal Tournament 2004, so we didn't place among the first five, but I will try again next year with a more advanced bot which implements the abilities of prediction of the opponent," says Carlos III University of Madrid professor Raul Arrabales. He says replicating human behavior in any environment is complex because it becomes necessary to combine different cognitive capabilities. Arrabales says the problem is that although much is known about the brain, it is only on a relatively high level, which prevents the creation of artificial neuron networks that mimic human ones with enough detail. The finalists in the competition were able to convince at least one of the judges that their bot was human.


Four NSF Grants Awarded to Stevens' Department of Electrical and Computer Engineering
Stevens Institute of Technology (10/08/09)

The National Science Foundation (NSF) has awarded Stevens Institute of Technology four grants that total more than $1 million to help fund studies that advance cognitive radio (CR) technology, which in the future could make networks faster and more secure. Stevens professor Rajarathnam Chandramouli will use an NSF grant to study methods of developing CRs that mirror human psychological and social abilities, including the ability to learn from circumstances and human behavior. Chandramouli says the team will study "how humans use or misuse their cognitive abilities to evolve into different societies. Our approach is inter-disciplinary, cutting across anthropology, drama theory, wireless networking, and stochastic analysis." Professor Yu-Dong Yao is using an NSF grant to study xBeam, a defensive program used to counter denial-of-service (DoS) attacks on wireless networks. Professor K.P. Subbalakshmi will use her NSF grant to study DoS attacks on dynamic spectrum access (DSA) networks. Professor Hongbin Li will use his NSF grant to find ways to create a united framework for wireless sensor nodes. "These techniques will afford improved awareness of the dynamically changing environment in a cognitive network," Li says. He hopes that his research will counteract issues with network bandwidth and power.


Recommender Systems Make Learning More Fun
Centre for Learning Sciences and Technology (10/13/09) Winnubst, Marga

As part of the European Union TENCompetence project, the Centre for Learning Sciences and Technologies' Hendrik Drachsler searched for the most suitable way to recommend learning activities, considering the personal needs, preferences, and learning goals of community members. Drachsler explored commercial navigation and recommendation systems and concluded that the specific behavior of community members required a combination of different recommendation techniques. Drachsler developed a prototype navigation system and ran experiments to determine if such a system offers additional value to informal learning communities. The project concluded that the system has a significantly positive influence on learning. The experiment group needed less time to complete the same number of learning activities, and participants in the group chose more personalized learning paths than members in a control group, primarily because they explored more learning paths. A second experiment showed there was no noticeable difference between informal and traditional learners, but that informal learners were more satisfied with the experience. The results of the study led to a second prototype, called ReMashed, which shows how recommendation systems can benefit from the information gathered through Web 2.0 technology by life-long learners in a learning network. ReMashed recommends the most suitable sources using tags and evaluations from individual learners.


Seeing Things
MIT News (10/13/09) Hardesty, Larry

Massachusetts Institute of Technology (MIT) professor Antonio Torralba and students from the school's Computer Science and Artificial Intelligence Lab (CSAIL) say they have developed an object recognition system that requires no training and can identify objects at least as well as and any other available program. The CSAIL system uses a modified version of a motion estimation algorithm. Consecutive frames of video normally change very little, so data compression schemes often store the unchanging elements of a scene once and only update the positions of moving objects, with motion estimation algorithms determining what objects move. The MIT system treats unrelated images as if they were consecutive frames in a video sequence. When the modified motion estimation algorithm attempts to determine what objects have "moved" between one image and the next, they system can identify objects of the same type. "It's a real commonsense solution to a fundamental problem in computer vision," says University of Central Florida professor Marshall Tappen. "The results are great and better than you can get with much more complicated methods."


Increased Success a 'Virtual' Certainty for Rugby Players
Queen's University Belfast (10/14/09)

Researchers from Queen's University Belfast are immersing rugby players in realistic simulated environments in an attempt to learn more about their perceptual skills and how they use visual information as they make decisions during games. As part of a new virtual reality training program, the rugby players wear a backpack of sensors and a head-mounted display for presenting a series of 360 degree virtual scenarios. "By presenting stereoscopic images in a head-mounted display and tracking head movements, the user's viewpoint is automatically updated giving a 360 degree virtual experience," says lead researcher Cathy Craig. PhD student Gareth Watson adds that "by controlling the events presented to the players, we can see how the visual information available to the participants at any moment in time influences the player's decision about when and how to act." Jeremy Davidson, forwards coach with Ulster Rugby, believes the virtual training scenarios can help develop a player's peripheral vision.


Dispute Finder: Making the Call on Web 'Facts'
Christian Science Monitor (10/13/09) Gaylord, Chris

Intel Labs researcher Rob Ennals has developed Dispute Finder, software that highlights inaccurate online information in pink with a link pointing toward a reliable body of evidence that disputes it. Dispute Finder, which currently only works with the Firefox Web browser, can be downloaded from disputefinder.cs.berkeley.edu. The program also has an ignore feature that, when activated, will prevent it from highlighting a certain factual error again. "I'm not trying to create a consensus or show the truth," Ennals says. "Something can be true and be disputed." He also wants to avoid opinions or biases, and thus relies on neutral source material that is not partisan. Dispute Finder relies on its 10,000 users to keep it abreast of new information. Ennals says that registered users can highlight disproven statements and label them as disputed. Conversely, they can mark a reliable source of information as evidence. Although he considers the program experimental, Ennals has already improved upon the software. Earlier this year, Dispute Finder could only monitor Web sites that its users had labeled themselves, but now it can highlight specific allegations wherever they appear on the Web. Ennals hopes that Dispute Finder will be expanded to social Web sites and Internet TV in the future. "So, if you're watching one of the more crazy cable news shows, a little box might appear that lets you know that the New York Times or Christian Science Monitor has disproved this," he says.


Louisiana Tech Receives DOE Grant for Cyberspace Education Program
Louisiana Tech University (10/05/09) Guerin, Dave

Louisiana Tech University has received a $951,000 grant from the U.S. Department of Education to support interdisciplinary cyberspace and science education initiatives in the state. The program, "Cyber K-12: Building a Foundation for Cyber Education in North Louisiana," is led by professor Galen Turner and will help train elementary through high school teachers in northern Louisiana. Cyber K-12 is one facet of Louisiana Tech's STEM Talent Expansion Program. Cyber K-12 also collaborates with the Cyber Innovation Center (CIC) to showcase the Cyber Discovery Summer Camp, which teaches educators and students about the history of cyber and its technological, social, and political benefits. "It shows students how life is interconnected and that they must pay attention to all of the issues surrounding the real problems that we face as a society," Turner says. Louisiana Tech's Les Guice hopes that Cyber K-12 will expand to become a national resource, serving as a model for universities in every state. "National leaders have emphasized the great importance in protecting the U.S. from cyberattacks," he says. Turner believes that Cyber K-12 will provide students with new opportunities and make cyber a more accessible discipline. "We're still teaching science, math, and engineering concepts; we're just doing it in a much more fun and dynamic way," says CIC's G.B. Cazes.


Abstract News © Copyright 2009 INFORMATION, INC.
Powered by Information, Inc.


To submit feedback about ACM TechNews, contact: [email protected]




Unsubscribe
Change your Email Address for TechNews (log into myACM)