Association for Computing Machinery
Welcome to the May 19, 2010 edition of ACM TechNews, providing timely information for IT professionals three times a week.


"Digital Genome" Safeguards Dying Data Formats
Reuters (05/19/10) Rhodes, Jason

European researchers have deposited a "digital genome" time capsule inside a data storage facility known as the Swiss Fort Knox, which contains a blueprint that future generations can use to read data stored using obsolete technology. The capsule is the result of the four-year Planets project, which was launched to preserve the world's digital assets as technology changes. "The time capsule being deposited inside Swiss Fort Knox contains the digital equivalent of the genetic code of different data formats," says British Library archivist Adam Farquhar. Planets project researchers note that the European Union alone loses at least three billion euros worth of digital information every year. "Unlike hieroglyphics carved in stone or ink on parchment, digital data has a shelf life of years, not millennia," says University of Technology of Vienna professor Andreas Rauber. The project aims to preserve data DNA, the information and tools to access and read historical digital material and prevent digital memory loss into the next century. "If we can nail the next 100 years, we figure we will be able to nail the next 100 years as well," Farquhar says.

HP Researchers Say Manure Could Help Power Data Centers
San Jose Mercury News (CA) (05/18/10) Bailey, Brandon

Hewlett-Packard researchers have proposed a biogas recovery system that converts livestock waste into methane, which can be used as fuel to generate electricity for data centers. The system would use the heat produced by the server banks to assist the process of converting the animal waste into fuel. Although many large data centers in the Pacific Northwest use hydroelectric power, HP scientist Chandrakant Patel says a data center should be able to tap multiple energy sources, depending on whether solar, wind, or biogas power is available, while using nonrenewable power as a backup. The researchers estimate that they would need 10,000 cows to power a 1-megawatt data center, which is a small to medium-size facility. Biogas power could also be used in developing nations, where electricity is expensive and the existing grid cannot support a large technology infrastructure. "This could be an opportunity for emerging economies where the need for [information technology] services will be greater and greater," Patel says.

Seeing Is Understanding--Using Artificial Intelligence to Analyse Multimedia Content
ICT Results (05/17/10)

The European Union-funded Bootstrapping Ontology Evolution with Multimedia Information Extraction (BOEMIE) project has developed artificial intelligence techniques to build and refine highly structured knowledge bases that can automatically identify, analyze, and index multimedia content. As BOEMIE analyzes a multimedia feed, it uses newly developed video, image, audio, and text analysis tools to extract as much information as it can. The BOEMIE Ontology Evolution Toolkit then tries to organize the information, and can go back online to search the Web for more data if necessary. After enriching the knowledge base, the system then re-analyzes the same footage, using the newly enriched ontology, which enables it to extract more information and propose additional refinements to the knowledge base. After testing the system, the researchers found BOEMIE enabled them to obtain information from multimedia sports coverage much more efficiently and accurately than existing automated systems.

Engineer Says Robotics Can Use a Woman's Touch
Womens eNews (05/17/10) Walton, Marsha

Texas A&M at College Station computer science professor Robin Murphy creates rescue robots designed to slither through collapsed buildings, fly over wildfires or floods, and check the integrity of a bridge from underwater, sending back live video and audio. "We've gone from things that look like a camera on wheels to things that look like an eight-foot-long caterpillar," Murphy says. After a disaster, a small flying robot could help in the long-range planning for survivors by providing information about the surrounding area and how people are using it. Murphy encourages other women to pursue rescue robotics research because the field is so new and could benefit from female-influenced research styles. She says the most effective rescue robot design is slow moving, painted yellow or orange, and has lights underneath so a victim can see it approaching.

Software That Learns by Watching
Technology Review (05/19/10) Graham-Rowe, Duncan

Massachusetts Institute of Technology (MIT) researchers have developed KarDo, tech support software that can automatically configure an email account, install a virus scanner, or setup a virtual private network. KarDo only needs to watch an administrator perform a task once before being able to carry out the same job on computers running different software. However, unlike macro programs, KarDo also attempts to learn the goal of each action in the sequence so it can be applied more generally later, says MIT's Hariharan Rahul. "I can go to my desktop, click on the Internet Explorer icon, go to a Web site, and then click on a particular link to download a file," says MIT professor Dina Katabi. The same actions could then be applied by KarDo on a machine running a different Web browser. The ultimate goal is for KarDo to intervene completely automatically, so that when a user sends a request to the tech department, KarDo would perform the task automatically.

Innovation: Teaching Robots Some Manners
New Scientist (05/17/10) Barras, Colin

People are more likely to adapt to and use robots if they behave more like humans, even if that means they operate less efficiently. For example, researchers at ATR laboratories in Japan recently tested how long of a response people preferred when giving commands to robots. Robots that are programmed to wait before completing a task are considered more welcome than those that respond immediately. "In the interaction with the robot, what is efficient is that the robot follows the norms of the conversation, which includes a seemingly inefficient one-second delay," says the University of Wisconsin-Madison's Bilge Mutlu. University of Washington researchers Peter Henry and Christian Vollmer are developing ways to help robots learn to move through a crowd as humans do. Instead of pre-programming fixed instructions, they believe it is easier to drop a robot untrained into the real world, but equip it with the ability to study and mimic the behavior of those around it. "If a human takes a geometrically longer route avoiding the crowd, our planner would learn to do the same thing," Henry says.
View Full Article - May Require Free Registration | Return to Headlines

Software Tool Helps Tap Into the Power of Graphics Processing
NCSU News (05/17/10) Shipman, Matt

North Carolina State University (NCSU) researchers have developed a compiler that could make it easier for traditional software programs to take advantage of graphics processing units (GPUs). "We have developed a software tool that takes computer program A and translates it into computer program B--which ultimately does the same thing program A does, but does it more efficiently on a GPU," says NCSU professor Huiyang Zhou. NCSU's compiler converts a program not optimized for use on GPUs into a program that can take advantage of the GPU's hardware to operate much more quickly. After the testing the new system, the researchers found that programs translated by the compiler ran approximately 30 percent faster than those optimized by leading GPU developers. "Tapping into your GPU can turn your personal computer into a supercomputer," Zhou says.

Giving New Meaning to 'Smart Car'
McMaster University (Canada) (05/17/10)

McMaster University computer and software engineers are studying how one IBM multicore processor can be used to connect a vehicle's automotive systems. Vehicles currently have multiple microprocessors that work in isolation, but integrating data from sensors and microprocessors has the potential to give cars the "cognitive" ability to avoid traffic accidents, find new routes when roads are congested, and predict when they will fail. "Investigating how a powerful multicore processor could be applied to manage that functionality will go a long way in helping build a smarter car that helps drivers operate their vehicles more safely and efficiently," says McMaster professor Alan Wassyng. The researchers also will study how an increase in computing power would help vehicles connect with regional and global transportation systems.

Feds Seek Game-Changing Cybersecurity R&D (05/13/10)

The White House has identified tailored trustworthy spaces, moving target, and cybereconomic incentives as themes for encouraging future game-changing research and development in cybersecurity for the federal government. In a notice recently published in the Federal Register, the National Coordination Office for the Networking and Information Technology Research and Development Program outlines the three themes and their challenges, including the high cost of securing information infrastructures, an asymmetrical cost of attack that favors attackers, and the lack of meaningful metrics and sound decision-making when allocating security resources. "Achieving enduring trustworthiness of the cyberspace requires new paradigms that re-balance security asymmetries of today's landscape," says the notice. The National Coordination Office is seeking public comment to help refine the themes. For example, the office wants to know how the themes could be enhanced, how organizations would support or incorporate the themes, and whether they can cite any state-of-the-art activities and use cases that support the themes.

Machines That Learn Better
MIT News (05/18/10) Hardesty, Larry

Massachusetts Institute of Technology (MIT) researchers have developed Church, a probabilistic programming language designed to cut the time it takes to build a machine-learning system to a matter of hours instead of months. Church is based on an inference algorithm, which instructs a machine-learning system on how to draw conclusions from the data presented to it. The algorithms currently used in probabilistic programming are designed to handle discrete data but struggle with continuous data. However, at the recent International Conference on Artificial Intelligence and Statistics, MIT student Daniel Roy presented a paper in which he and MIT instructor Cameron Freer describe an inference algorithm that can handle large classes of problems involving continuous data. Their work could be especially useful for artificial intelligence systems whose future behavior is dependent on their past behavior, says Rutgers University computer scientist Chung-chieh Shan.

Berkeley Lab Scientists Build Software Framework for ATLAS Collaboration
Lawrence Berkeley National Laboratory (05/17/10) Vu, Linda

When Lawrence Berkeley National Laboratory's (LBNL's) ATLAS experiment is turned on, its detectors record about 400 proton collision events per second, a rate equivalent to filling 27 compact discs per minute, in an attempt to find new dimensions in space, undiscovered physics forces, and the origins of mass. Three thousand researchers in 37 countries must process and analyze all of this data and collaborate on results in real time. To facilitate this distributed workflow, the researchers are relying on a software framework called Athena, which was developed by an international team of scientists led by LBNL's Paolo Calafiura. "When developers plug their codes into the Athena framework, they get the most common functionality and communication among the different components of the experiment," Calafiura says. Athena enables scientists to focus on developing tools for analysis and analyzing data, instead of worrying about infrastructure issues like the compatibility of files. Athena is built on top of the Gaudi software framework, which was originally developed for the LHCb experiment.

Virtual Reality Used to Transfer Men's Minds Into a Woman's Body
Guardian (UK) (05/12/10) Sample, Ian

Barcelona University scientists recently conducted an experiment in which they used virtual reality to transfer men's minds into a woman's body. In the study, men wore a virtual reality headset that enabled them to see and hear the world as a female character, even changing how they saw their own body and clothes. The researchers say the body-swapping effect was so convincing that the men's sense of self was transferred into the virtual woman, causing them to react reflexively to events in the virtual world in which they were immersed. The men reported feeling as though they occupied the woman's body and even gasped and flinched when the woman was slapped by another character in the virtual world. "This work opens up another avenue for virtual reality, which is not just to transform your sense of place, but also your sense of self," says University College London's Mel Slater.

Enabling Video Systems to React Intelligently to Content
EUREKA (05/12/10)

The EUREKA ITEA 2 software Cluster CANTATA project has yielded methods for the automatic analysis of digital video content for various applications, including enhanced surveillance, expedited medical diagnosis, and fast consumer access to home entertainment material. The project has imbued within the video system awareness of both content and its context, and the ability to apply this awareness toward the establishment of an action or autonomous environmental control. Among the project's innovations are algorithms for content analysis in different domains, a universally appropriate analytical and presentation platform, visualization and user interaction with a core concentration on home consumers, and techniques for validation of content-aware products. CANTATA produced a content-aware interactive TV system that supports many new features that deploy multimedia content awareness, such as the automatic summarization of broadcast news and sports items. Other CANTATA innovations include an intelligent surveillance camera that blends together advanced video-content analysis and video compression for streaming content over the Internet.

SARA Opens Gate for HPC Cloud Researchers
HPC Wire (05/10/10) Hemsoth, Nicole

The Netherlands' BiG Grid national grid project's SARA national high-performance computing (HPC) and e-Science support center has established an experimental HPC cloud environment that will offer Dutch researchers an opportunity to operate within their own virtual private HPC cluster. The cluster is capable of hosting individual configurations that will function according to the needs of each scientific team and scale accordingly. The SARA team is especially interested in applications which are difficult or nearly impossible to run on the existing HPC platforms but do run in a local environment. The success of the experiment could end up being a key proof of concept for members of other scientific HPC communities worldwide. SARA cloud project leaders Tom Visser and Floris Sluiter note that "because the HPC infrastructure and computing environment is fully configurable on demand to specific needs, the user can save time and effort porting their applications to a specific HPC platform." Visser and Sluiter believe SARA will help accelerate the cycle from scientific question to computational solution.

Abstract News © Copyright 2010 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