Dozens of Johns Hopkins University students enrolled in a computer security course last semester learned how painfully cheap and easy it is to acquire personal data online when they were grouped into teams assigned to aggregate, clean, and link entire databases of dossiers on Baltimore citizens using only public data sources with a maximum budget of $50. Several teams collected upwards of 1 million records on hundreds of thousands of individuals. The project was the brainchild of Johns Hopkins computer science professor Aviel Rubin, who is also technical director of the university's Information Security Institute. Some participants obtained information by filing Freedom of Information Act requests at local government offices, while others tapped whole databases from online sources or free commercial address databases using special computer scripts. Profiled citizen David Albright was troubled by how effortlessly information such as his occupation, address, phone number, birth date, and party registration was gathered: "What would be disturbing is if by having all this information consolidated, it made stealing an identity easier," he said. Privacy proponents have similar concerns, especially in regards to how easy it is to access Social Security numbers. ACLU lawyer Jason Brandeis expressed the need to balance out the protection of individual privacy and the public interest in unfettered access to government data. Rubin concluded that "there are strong negative consequences to being able to collect and correlate all this information on people, but it is also possible that the consequences to personal freedom would be worse if it were outlawed."
Click Here to View Full Article
(Articles published within 7 days can be accessed free of charge on this site. After 7 days, a pay-per-article option is available. First-time visitors will need to register.)

The latest proposal from the European Parliament to amend the European Union's software patents directive has provoked complaints from software and technology groups claiming the revisions would violate the essentials of existing European patent law. The proposal contains 40 suggested amendments from Socialist parliament member and former French prime minister Michel Rocard, who is calling for a ban on all software patents that do not control a physical process--specifically, a "controllable force of nature." Eicta President Mark MacGann says the amendments would make the granting of patents in areas such as data/video/audio compression, data processing, encryption, and speech coding impossible. The Green members of parliament support Rocard's proposal, with Austrian Green MEP Eva Lichtenberger stressing that software patents hurt small and midsized companies by putting most patents in the hands of a few major companies. Her feelings are shared by many other deputies, who believe such patents choke innovation. Parliament has until July to vote on the proposed revisions, and many MEPs are confident the parliament will support a less industry-friendly regime than the one member states endorsed earlier this year.
Click Here to View Full Article

Videoconferencing technology has made considerable strides, but cost and quality issues have hindered its mainstream penetration. University of California at Berkeley researchers have developed the MultiView videoconferencing system, a tool that creates a "shared space" feeling among users and supports nonverbal communication, according to UC Berkeley researcher David Nguyen; in addition, the system is built from relatively cheap materials. MultiView employs multiple video streams, each displayed at a different location, while the retroreflective materials that make up the multi-viewpoint screen allow light from the projector in front of each conference participant to be reflected toward that person. Cameras are also placed to maintain geometric relationships during a group teleconference; every group member has a camera positioned directly above the image of the person seated opposite them in the remote group, allowing attendees to receive novel perspectives aligned to the camera locations. "In our testing, we have been able to show that not only are members able to tell when a remote person is looking or not looking at [him, the person is] able to tell with a high degree of accuracy who they are looking at specifically," says Nguyen. MultiView sets up a virtual conference room with a table where the participating groups sit opposite each other, and Nguyen says the system facilitates natural nonverbal exchanges based on gesture, direction of gaze, posture, proximity, and gaze awareness. Nguyen says the system, which was detailed at the Computer-Human Interaction conference (CHI 2005) in April, is ready for practical use.
Click Here to View Full Article

The Real ID Act is causing serious concern among state technology officials, who are unsure about what is required and their ability to implement necessary changes. Delaware CIO and NASCIO President Thomas Jarrett says that while exact technical requirements are unclear, Real ID will undoubtedly involve some uncomfortable changes, especially since it requires states to share information. "As technologists, we start getting paranoid when we move information outside our walls," he says. Real ID was passed by Congress and approved by President Bush last week and is meant to improve homeland security by standardizing identification requirements. Privacy advocates say the bill amounts to a federal ID law since state residents will be unable to use driver's licenses for federal functions, such as boarding an airplane, unless those cards comply with Real ID rules. Those requirements include making cards machine readable and checking with other government databases before issuing compliant cards; the deadline for the new system is 2008. Gartner research director Richard Hunter says the technology aspect will not be as challenging as mustering the resources necessary to restructure databases and coming up with rules for access, management, and security for shared records. EDI links are already shared among 39 states to verify driver's license applicants' names and Social Security numbers, but Real ID extends documentation requirements to include birth certificates and other vital information. Systems will have to be built to store those records digitally, and the Electronic Verification of Vital Events pilot project is one possibility that has already proven itself. But American Association of Motor Vehicle Administrators CIO Jay Maxwell says document-checking systems will only be effective after many years' worth of data has been populated into the system.
Click Here to View Full Article

For information regarding ACM stand on the Real ID Act, visit http://www.acm.org/usacm.

Open-source SPIRAL software tools developed by U.S. university researchers with funding from the National Science Foundation and the Defense Advanced Research Projects Agency could radically change the writing of computer code, especially when measured against the latest breakthroughs in high-performance hardware that is frequently running Linux. Carnegie Mellon University professor Jose Moura says the software can shave time off the testing and debugging process by automatically generating less buggy code. He says it is typical of hardware manufacturers to develop libraries of algorithms to be employed in programs--libraries that are used by programmers to process data and solve equations; but the library development process can be unwieldy. Essentially, the SPIRAL tools represent an extensive set of the algorithms using minuscule constructs, which are used to organize the algorithms according to rules. Moura says the SPIRAL development team customized the software for IBM's Blue Gene/L supercomputer so it could function as a quality demonstration of both the machine and SPIRAL applications. Blue Gene systems architect Jose Moreira says SPIRAL was used to generate the fastest FFT library for the supercomputer. "The fact that SPIRAL uses an automated approach to code optimization results in scientific libraries that can be highly optimized to each specific architecture, including Blue Gene/L," he notes.
Click Here to View Full Article

Computer and video games are gaining credence in academic institutions as an area of study, despite skepticism from university administrators and traditional computer scientists. Seattle's DigiPen University was the first U.S. school to offer a four-year game development degree, while a few more four-year universities in the country started offering undergraduate majors in game development last year. There is enthusiasm among academics to study game and gamers' behavior from a cultural or anthropological perspective, although several professors promised to disassociate themselves from the practical side of the gaming industry at a recent conference discussion. Researchers at the University of Wisconsin, Madison, mainly investigate how games can enhance learning, while other researchers focus on a range of topics that include gaming sociology and demographics and the economics of multiplayer games; however, programs focusing on the more practical aspects of gaming are starting to thrive. Curriculum is usually divided into programming, design, and art categories, with students permitted to concentrate in only one of these areas at some schools and in all three at others. Commercial software-development kits are often too costly for schools, so many students work on PCs instead. University of Denver professor Scott Leutenegger says he warns incoming students that the gaming sector "is an industry with high burnout rates, long hours and incredibly tight deadlines," and he expects most students will want to change careers within a few years.
Click Here to View Full Article

Sifting through large volumes of data produced by high-energy physics experiments and other research projects for specific information just became easier thanks to the Word-Aligned Hybrid (WAH) compression technique developed and patented by researchers at the Energy Department's Lawrence Berkeley National Laboratory. WAH is used in FastBit software to compress bitmap indexes and optimize them in terms of computational complexity, and tests demonstrate that WAH-compressed indexes run substantially faster searches than other bitmap indexing schemes. "What makes our compressed bitmap index really special is that it is not only theoretically optimal but also practically more efficient than any other indexing scheme tested," asserts John Wu of Berkeley Lab's Scientific Data Management Research Group. Other Energy Department research projects currently employing WAH technology include an initiative for tracking ignition kernels in a hydrogen-air combustion simulation; the Dexterous Data Explorer project, which uses FastBit to visualize large scientific datasets based on queries; and view-dependent software that displays isosurfaces of large complex data in real time. Other institutions where FastBit could be useful include CERN, which is considering fully integrating FastBit with ROOT object-oriented data analysis framework software. This would open up FastBit's search capability to many users, given the employment of ROOT software in most high-energy physics projects around the world. The International Supercomputer Conference recently selected a paper detailing the WAH method and its application as a "best paper," and the document will be presented at the conference in late June.
Click Here to View Full Article

Carnegie Mellon University computer scientists Seth Goldstein and Todd Mowry have conceived of shape-changing robots that can assemble themselves into replicas of human beings or other objects that can sense, change color, and move in three dimensions. Each individual unit or claytronic atom (catom) would be studded with electromagnets to affix itself to other catoms, and move by rolling itself over other catoms; the catoms would change color using light-emitting diodes and sense light through photo cells, and each robot would be equipped with a Pentium-class computer. Mowry says a moving shape such as a human replica could be comprised of 1 billion catoms, so a rapidly reconfigurable computer network architecture is called for. Catoms may identify themselves by their individual position or function, since identification by number is probably impractical. Jason Campbell with Intel Research Pittsburgh says the large number of catoms requires a scheme that compensates for the unavoidable failure of individual robots, while Goldstein says power issues can be addressed by having the catoms automatically organize themselves into electrical circuits. Campbell also notes that power interruptions are an inevitability, since connections will be repeatedly made and broken as the shape moves and the catoms reconfigure themselves; these interruptions can be compensated by outfitting the robots with a capacitor or small battery. Carnegie Mellon roboticist Metin Sitti says the catoms would be able to retain their shape even when the system is deactivated if their sides are covered by artificial fibers similar to the gripping foot hairs of a gecko, and adds that fabricating the fibers out of carbon nanotubes could allow them to double as electrical connections between catoms. The project has received funding from the Defense Advanced Research Projects Agency and the National Science Foundation.
Click Here to View Full Article

Mobile phones risk becoming increasingly difficult to use as manufacturers and carriers promote devices with more and more functions, said mobile industry design consultant Scott Jenson at a Microsoft Research conference held in Cambridge, United Kingdom discussing simplifying computing. The latest fad is music capability, with Nokia, SonyEricsson, and Samsung all including music functions on their new phone models; but the lukewarm reception to phones with integrated cameras and the convergence of more functions into one device presents major design challenges. Jenson says the industry needs to focus more on simplicity through the design, citing the iPod as a good example of how design can dramatically influence an industry: The iPod's scroll function is critical to making the device usable because it allows people to accelerate their track search by scrolling faster; the ability to go song-by-song or quickly scroll to the end of a 3,000-song list is taken for granted by the user, which is a hallmark of good interface design. Jenson said he is currently working for a mobile phone company to help build an integrated phone and music device, and that focus groups with young users showed iPods were often run until the battery was completely exhausted. This presents a problem for integrated phone devices, which should ensure users can place calls. Jenson is optimistic that new technologies such as multifunctional rubber conductive keys and rollable electronic paper screens will increase usability.
Click Here to View Full Article

The Touchy Internet system developed by researchers at the University of Singapore's Mixed Reality Lab enables users to feel a chicken remotely by stroking a chicken-shaped doll that moves in concert with a real chicken monitored by a webcam. Tactile data captured by the replica's touch sensors is sent by radio to a nearby PC, which in turn transmits the data to a remote computer near the live chicken via the Internet. The second computer activates vibration motors in a haptic jacket worn by the chicken, enabling it to sense the user's touch. Remote haptic interaction could enable zoo visitors to touch animals without putting themselves at risk, while people with allergies to dogs and cats would be able to physically interact with their pets. Further advancement of the technology could enable remote haptic interaction between human beings, and the team that developed Touchy Internet plans to create a haptic suit for the purpose of "Internet hugging." Team member Ling Shang Ping says the technology could be applied to remote dance synchronization, in which people wearing sensor-studded shoes can learn how to dance by having their muscles stimulated in tandem with the dance teacher's movements. Director of the University of Southern California's Information Laboratory Cyrus Shahabi expects at least 10 years to pass before an object's texture can be haptically reproduced with high fidelity. He also says current haptic devices cost too much to have consumer appeal.
Click Here to View Full Article

Despite a number of false starts over the last decade, 3D interfaces for the Web are ready for widespread use, says Virtual Reality Modeling Language (VRML) co-creator Tony Parisi. VRML was first deployed commercially 10 years ago, but for different reasons Web 3D technology never took off as predicted, and even efforts from the likes of Microsoft, Adobe, Intel, and Macromedia failed to garner substantial momentum. Parisi says the factors needed for widespread use have finally come together: Interest is picking up from individual developers, VRML has been adopted by the Web3D Consortium as an XML-based ISO-approved standard called X3D, PCs have requisite hardware and broadband connections, and applications are starting to break away from the standard Web browser platform. In addition, new art school graduates are trained in 3D development tools, similar to how Photoshop experts helped power the Web explosion earlier. What makes the current Web 3D applications different from previous ones is that they are not merely technical demonstrations but actual applications. Gaming applications represent the cutting-edge of 3D interfaces, especially 3D gaming universes such as Linden Labs' Second Life, which Parisi compares to Prodigy and CompuServe at the beginning of Web commercialization. He also says companies making use of the open X3D standard are similar to Netscape, which popularized the Web with its browser. Web 3D technology has far more broad implications than merely more realistic gaming: Parisi says 3D interfaces will organize information in much more effective ways, with the feel of a TV broadcast, for example.
Click Here to View Full Article

The Defense Advanced Research Projects Agency (DARPA) is committing approximately $500,000 to BAE and three MIT professors to create software that could model the conditions under which a nation state fails and slips into chaos. Washington's interest in such a tool partly stems from concerns that lawless countries are ideal terrorist havens. MIT professor and DARPA grant recipient Nazli Choucri directs the university's Technology and Development and Middle East programs. Fellow recipient Stuart Madnick of the Sloan School of Management has spent his career developing technologies that can aggregate and analyze data from a broad spectrum of sources. Corporations use Madnick's research to help process data that can be harnessed for the pursuit of profitability. William Zartman, head of Johns Hopkins University's conflict management program, doubts that software can take reliable measurements of a troubled nation's goings-on. He wonders how understated changes in political alliances can be incorporated into a computer model, and argues that spies and scholars are much more dependable sources of intelligence.
Click Here to View Full Article

University of Houston computer science professor Ioannis Pavlidis, with the help of his Infrared Imaging Group at the College of Natural Sciences and Mathematics, has developed the Automatic THErmal Monitoring System (ATHEMOS) that can physiologically monitor human users without touch. ATHEMOS is part of Pavlidis' effort to revolutionize human-computer interaction by making computers aware of their users' emotional and physical states in order to facilitate more appropriate responses. The ultimate goal of Pavlidis' "Interacting with Human Physiology" research is to tap computing resources at home and in the office and integrate them with new sensing, algorithmic, and interface techniques to augment user experience while simultaneously constructing a new model for preventive medicine. The initiative calls for a computer equipped with a thermal imaging camera that functions as a peripheral, modeling facial imagery by bioheat to extract vital signs and make deductions about numerous health symptoms continuously. The National Science Foundation's Division of Information and Intelligence Systems recently awarded a three-year, $640,169 research grant to Pavlidis' project. The project's human experimentation angle will involve the professor's collaboration with Columbia University's Medical Usability Lab and the Physiology Lab of the Mayo Clinic in Rochester, Minn.
Click Here to View Full Article

Researchers in Virginia Tech's Bradley Department of Electrical and Computer Engineering will explore ways to meld mobile ad hoc networks and wireless sensor networks by developing a testbed platform with a $246,000 Defense University Research Instrumentation Program (DURIP) grant from the U.S. Defense Department. Making these two networks interoperable is an important step in the creation of a "network-centric" communications infrastructure, which the DoD views as a crucial element in future military operations, according to DURIP project principal investigator Thomas Hou. Mobile ad hoc wireless networks can establish links between military groups that need to maintain communications while traveling, while wireless sensor networks can be deployed in hostile areas to relay observations to personnel out of harm's way. "We plan to construct a two-tiered logical network architecture with a wireless sensor network on the lower tier and a mobile ad hoc network on the upper tier," says Hou. "This architecture should seamlessly integrate the sensing capabilities of the sensor network with the processing and communications capabilities of the ad hoc network, all within a common platform." James Thorp, director of electrical and computer engineering, says the testbed platform will substantially improve Virginia Tech's ability to carry out future DoD-funded research, as well as provide better wireless communications and networking courses to graduates.
Click Here to View Full Article

As network and general infrastructure security improves, hackers will increasingly target line-of-business applications. Enterprise applications are increasingly vulnerable for three reasons: Growing demands for application integration between systems that were never designed to work together; costs for securing enterprise applications fall upon a more narrow segment of users than with general platforms; and new supply chain demands require applications accommodate more users and be flexible. As a result of these pressures, enterprise application developers are faced with vast stores of code that could potentially be abused by a hacker. Firewalls, stronger authentication, and greater documentation will not help protect these vulnerabilities since the applications are completely exposed to authorized users; the trend toward greater abstraction in business application development does not help either, because abstraction conceals details from developers that hackers will look for. The only sure way to secure enterprise applications is to build security from the inside out, such as by writing validation rules and bounding code with exception handling so it will only perform as intended. Anitian consultant Andrew Plato says many enterprise application developers are totally oblivious to hacker threats, but need to be educated on network and infrastructure security as new business applications are network-facing. The Sarbanes-Oxley Act could help strengthen business application development as it requires the characterization of system function and control and is also causing companies to separate coding and deployment tasks. IT governance tools such as those used at the Department of Defense could also help mainstream IT managers lock down their enterprise application development.
Click Here to View Full Article

The National Science Foundation's decision to withdraw funding for its three supercomputer centers is breeding uncertainty about the future of academic supercomputing in the United States. NSF supercomputers are critical to academic efforts because other federal supercomputing resources--though more powerful than NSF and university machines--are generally inaccessible to academics. The National Coordination Office for Information Technology Research and Development reports the foundation has requested a $307 million supercomputing budget for fiscal 2006, but this is paltry compared to the $1 billion-plus "cyberinfrastructure" budget recommended by an NSF advisory panel two years ago. The foundation elected to allow the five-year contracts for the San Diego Supercomputer Center, the National Center for Supercomputing Applications, and the Pittsburgh Supercomputing Center to expire. The agency has supplied additional money to sustain the operation of the first two facilities for three more years, and intends to invite new bids for their operation to researchers; but there is no assurance that any of the centers will get a contract, nor is any money being budgeted for new computer purchases. Researchers warn that the centers' host institutions stand to lose hundreds of millions of dollars in outside investment if the facilities are transferred, and Russ Miller with the State University of New York at Buffalo says the funding uncertainty adversely affects employee morale. NSF officials argue that competition helps guarantee the facilities fulfill scholars' requirements and are a good investment of taxpayers' money, while the foundation's Sangtae Kim is confident the centers can survive on grants from the NSF and other agencies.
Click Here to View Full Article
(Access to this article is available to paid subscribers only.)

The integration of relational database technologies with an iterative software development process is a critical element in many software projects, and achieving this task requires cooperation between developers and data professionals. "Agile Database Techniques" author Scott Ambler advises the latter group to embrace evolutionary methods similar to those used by the former. He cites a series of columns outlining an evolutionary data modeling strategy called Agile Model Driven Development (AMDD): The technique involves creating a slim conceptual domain model representing the main business entities and their relationships to one another, which is then used to direct physical class and modeling initiatives during development iterations on a just-in-time basis. Benefits of AMDD include a minimization of waste, evasion of major rework, reduction of the modeling effort in general, and simplification of object/relational mapping. Ambler also says data professionals must familiarize themselves with database refactoring, a job complicated by "the prevalence of increased coupling." He notes that some database refactorings are easier to deploy than others, and recommends that both code and database schemas be run simultaneously during a transition interval long enough to allow the other project teams to update and implement their applications. Enabling both types of refactoring requires the presence of a regression test suite, separate work areas, solid tools, and a configuration management category. Ambler lists several steps to boost agility even further, such as enabling developers, data professionals, and business stakeholders to work close together every day; sharing skills and learning new skills from others; exploiting enterprise assets and standards collaboratively; never working alone; and striving to decrease the feedback cycle.
Click Here to View Full Article

The human brain is superior to the computer in terms of operational efficiency and functions such as vision, hearing, pattern recognition, and learning; the key to this superiority appears to lie in the organization of the brain's neural system, which engineers are attempting to duplicate electronically. Such a breakthrough could yield implantable electronic retinas that restore sight or sound processors that restore hearing, as well as smart visual, audio, or olfactory sensors for robots, writes University of Pennsylvania bioengineering professor Kwabena Boahen. Power-efficient microchips patterned after the neural system could form the basis of such advanced technologies, and University of Pennsylvania researcher Kareem Zaghloul has created a silicon retina that is 1,000 times less power-hungry than a PC. His Visio1 chip uses four types of silicon ganglion cells that mimic the way in which voltage-activated ion channels induce biological ganglia to discharge spikes. Boahen says hardware customization is common to both the brain and neuromorphic chips, and morphing the customization mechanism would make reverse-engineering the brain's circuits unnecessary. Research into neural development led to the realization that sensory neurons wire themselves in response to sensory inputs, and accept signals from neurons that are consistently active when they are active. This process was morphed, imperfectly, into the Neurotrope1 artificial tectum chip, and Boahen's team reasoned the system could perhaps be refined through deeper investigation of cortical connections. Researchers have successfully emulated in silicon the visual cortex's process of responding preferentially to object edges of a certain orientation, but Boahen notes that integrated circuits with many more transistors per unit area are needed if all six cortex layers are to be morphed.