The 9/11 Public Discourse Project has charged that the government is making insufficient progress in defending the Internet, communications networks, and other critical elements of infrastructure. The bipartisan organization, composed of former members of the Sept. 11 commission, gave a D rating to protection initiatives to shore up critical infrastructure, criticizing the government for not having conducted vulnerability assessments, defined national priorities, or recommended how the scant funding available should be dispersed. "Many obvious steps that the American people assume have been completed have not been. Our leadership is distracted," according to a statement issued by the project's leaders. They report that no key decisions are likely to be made in the next year, and described the government's failure as "shocking" and "scandalous." Project members also said the government must improve its ability to share information among agencies, and criticized the slow pace of implementing the Real ID Act, which calls for more secure identification cards. The system for scrutinizing foreign visitors also drew fire from the project members, who claimed that while the system is operating, some borders are still unsecured.
Click Here to View Full Article

The international team of researchers that won the Bandwidth Challenge at last month's SC05 broke the world record for network speeds with an official peak of 131.6 Gbps. In addition to their ability to sustain data throughput, entries in the Bandwidth Challenge were judged by their contribution to supporting science. The team was led by the California Institute of Technology, which took the top spot for the third consecutive year, transferring 475 TB of physics data from institutions around the world over a span of 24 hours, and sustaining average throughputs in excess of 100 Gbps for several hours. "This is much more than measuring how quickly we can send bits," said Don Patravick of the Fermi National Accelerator Laboratory, which partnered in the project. Patravick believes the project is especially significant for transferring actual physics data drawn from real experiments in an extensively collaborative effort. The Fermilab and the Stanford Linear Accelerator Center helped develop software to facilitate the data transfer, which took place over seven 10 GB optical links. The transfer of data over international links at such high speeds will advance data-intensive computing, as it helped shed light on the shortcomings of current processes. Data-intensive scientific applications will reach new heights when the Large Hadron Collider (LHC) starts to operate in 2007, making vast quantities of data about the origins of the universe available to thousands of scientists around the world across high-speed optical networks. The winning entry at the SC05 can be seen as a preview for the LHC, which is generally acknowledged to be the next generation of data-intensive scientific research.
Click Here to View Full Article

A team of Stanford computer scientists has developed an algorithm that will improve machine vision by enabling robots to gauge distances from a singular image. Poor depth perception is the principal reason robots have such difficulty maneuvering, though their ability has improved considerably with larger sensor arrays. Not every robotics project has the budget to install an expansive cluster of sensors, however, but it was previously thought impossible to derive "depth estimation from a single monocular image," said Andrew Ng, an assistant professor of computer science who led the research. In a basic version of Ng's algorithm, a radio-controlled car navigated through a densely wooded area for several minutes before crashing. The software enables the robot to glean from still images indications of depth, such as different textures, converging lines, and varying levels of haze. The software divides an image into sections, analyzing each individually and comparatively to detect variations. It also examines each section at different levels of magnification. Trials of the software found that robots misjudged depth by an average of roughly 35 percent, though Ng believes they will be able to adjust for this uncertainty. Robots using Ng's algorithm can also perceive obstacles between five times and 10 times farther away than conventional stereo vision algorithms that attempt to triangulate depth with two cameras.
Click Here to View Full Article

The University of New South Wales has used an open source license to create a programming language for schoolchildren based on Robolab, the language used to program Lego robots. Professor Paul Compton, head of UNSW's School of Computer Science and Engineering, believes students will find the programming language to be enjoyable and stimulating. In developing the simple icon-based language, Thomas Legowo, a 22-year-old computer engineering student, focused on improving the features of Robolab and making it easier to use. "You can use the icons to program originally," says Legowo. "Then you can look at the Not Quite C code automatically produced from your iconic program to see what is happening." Legowo also notes that the graphic programming language for children 8-15 years of age does not allow users to take any illegal steps, and includes a layout manager that always produces a fine layout. The program should be available in February.
Click Here to View Full Article

A group of researchers from Virginia Tech's business and engineering departments is developing technology to safeguard the privacy of children on the Internet. Children are often prompted for their personal information when using the Internet for games, interactive learning, or Web surfing. "While kids today are adept at using computer technology, most are still very naive about privacy protection. The promise of a small prize can easily convince them to share personal information," said Janine Hiller, a professor of business law. The Children's Online Privacy Protection Act of 2000 mandates that Web sites must obtain "verifiable parental consent" before soliciting information from children under 13, though enforcement has been plagued by the absence of technology to prove that consent actually comes from a child's parents. To get around this problem, the researchers developed POCKET (Parental Online Consent for Kids' Electronic Transactions), a system for obtaining verifiable consent without requiring direct supervision by the parents. The POCKET system allows parents to customize a disclosure policy for their child's information, offering a previously unavailable level of flexibility, and monitors the Web site's handling of the child's information through contract and log files. By making the system user-friendly, the developers increased the likelihood that it will be adopted on a large scale. The NSF has awarded the researchers a $450,000 grant to develop a prototype, conduct research about the obstacles to implementing privacy protections, and survey parents' views on children sharing information online.
Click Here to View Full Article

Digital forensics has emerged as a growth industry to combat the ease with which skilled users can manipulate and alter images with software programs such as Photoshop. Recent cases of newspapers doctoring photos, coupled with the concern that digitally altered images could be used as court evidence, led Polytechnic University computer scientist Nasir Memon to join the effort to expose digital alterations. One method of detecting tampering, known as digital watermarking, adds identifying data to an image, which is corrupted when the image is altered in any way, though the process is costly and not widely practiced. Because not every photo submitted in court has a digital watermark, digital forensics can be used to determine if an image has been altered and then match it to the camera used to take the picture. Memon developed software to identify a camera's manufacturer by identifying the company's unique interpolation algorithm, the tools which cameras use to alter colors to compensate for the red, blue, or green sensors that were not functioning in a given pixel at the time the picture was taken. Memon reports that his method is 90 percent accurate, and that he has defined the interpolation algorithms of 10 camera manufacturers. A more specific application, developed by Jessica Fridrich of the State University of New York, can match a photo to a specific camera through zooming in on the unique imperfections, known as noise, every camera produces. "We have discovered the equivalent of matching a bullet to the barrel and gun," she said. The software, which Fridrich claims to be 99.9 percent accurate, can still match a photo even if its file has been shrunk, unlike Memon's technique, which fails when a file is compressed.
Click Here to View Full Article

A recent report alleging that fair use rights are imperiled by vague copyright laws has called for a six-point program for reform, including such measures as moderating infringement penalties and increasing the pool of pro bono lawyers available to represent those accused of violating fair use laws. Fair use is designed to permit the copying of materials without the author's consent, so long as the purpose relates to "criticism, comment, news reporting, scholarship, or research," as stated in the longstanding statute. Marjorie Heins, coordinator of the New York University Law School's Brennan Center for Justice, which authored the report, believes that those terms are too ambiguous. The report polled stakeholders and reviewed letters sent to accused copyright violators, and found considerable confusion over the definition of fair use, as well as numerous infringement claims Heins felt were bogus, such as a cease and desist letter LucasFilm sent to a man operating a personal site called Tatooine.com, which contained no material relating to the Star Wars movies from which he borrowed the name. The report also criticized takedown notices which, sanctioned by the 1998 Digital Millennium Copyright Act, allow copyright owners to pressure ISPs and search engines to take down sites containing trademarked content, often without formal litigation. Fair use has been the subject of much congressional debate in recent years, where a minority voice calling for moderating copyright protections has been drowned out by protests that such moves would be tantamount to sanctioning piracy.
Click Here to View Full Article

A group of researchers from Carnegie Mellon University and Taiwan's Industrial Technology Research Institute (ITRI) are due to unveil a new software security toolbox with a host of capabilities, such as conducting surveillance at transportation centers and alerting parents to a teenager using their car without permission. ICTrack is built from modules, each representing a mathematical algorithm, that are flexible enough to track almost anything. An expansive software library powers the system, allowing users to design a custom security system on their own computer, and to supplement it with commercially available video equipment. The upcoming symposium in Taiwan will also see the unveiling of a motion sensor roughly the size of a dime, capable of tracking lost or stolen devices, such as a cell phone or a laptop. The sensor, which can be embedded in any commercial device, detects acceleration and sends an encoded signal, and can also be used as a navigational aid. Law enforcement officials have expressed interest in the device for tracking missing children, said Tsuhan Chen, an electrical and computer engineering professor at Carnegie Mellon. The symposium will also be a chance for Carnegie Mellon President Jared Cohon to promote the ITRI lab, which is providing $6 million over the next four years to ongoing research in various computing technologies.
Click Here to View Full Article

The Fabrication Labs that MIT has set up around the world are allowing people to use technology to solve their own problems. Makeda Stephenson, 13, decided to use a teen learning program at a Fab Lab to create a flight simulator program because her local computer stores did not offer any flight simulator games that she liked. MIT has set up seven Fab Labs, each with about $25,000 worth of computers and computer-controlled manufacturing tools to enable people in small communities to design and make their own technology applications. While herders in Norway have used a Fab Lab to build a telecommunications network for tracking the movements of sheep with radio antennas and electronic tags, villagers in Ghana have used the tech tools to harness solar power for electricity. More whimsical projects involve a teenage girl in Boston developing a security system that photographs anyone who goes near her diary, and a dress that spikes outward when someone invades the personal space of the wearer. Neil Gershenfeld, an MIT physicist and computer scientist, suggests Fab Labs could lead to a revolution in do-it-yourself invention.
Click Here to View Full Article

Cornell University and nine other universities worldwide will get significant funding from Microsoft to pursue high-performance computing (HPC) research. Microsoft announced the new grants at the recent international Supercomputing 2005 conference. Cornell was one of the 10 Institutes for High-Performance Computing selected by Microsoft to receive funding. The others were: the University of Tennessee, the University of Utah, the University of Virginia, Nizhni Novgorod University in Russia, Shanghai Jiao Tong University in China, Southampton University in England, Tokyo Institute of Technology in Japan, and the University of Stuttgart in Germany. Five years ago Cornell researchers demonstrated to Microsoft that programs needing a supercomputer's power could run under Windows. Last summer, Microsoft chief technical officer Ray Ozzie visited the Cornell's HPC hub, the Cornell Theory Center (CTC), and asked for project proposals. CTC executive director Linda Callahan says Ozzie "was extremely impressed with what he heard. He mentioned that we were doing research he hadn't heard other universities address." Cornell ultimately submitted four proposals, including providing easier access to genomic databases, using a Web interface to geographically distribute software components of an engineering simulation, and providing scholarly information in a Microsoft Office format. Microsoft's John Borozan says, "CTC is helping a number of open source bio applications reach compatibility with the Windows platform, and submitted Windows benchmarks for consideration and placement on the Top500 list."
Click Here to View Full Article

U.S. Rep. Mike Thompson (D-Calif.) believes lawmakers should set an example for the rest of the nation in recycling computers and other electronic waste. In November, he proposed a bill that would provide a standard for how Congress gets rid of its electronic devices. The House is expected to address his bill and proposal to reduce electronic waste in early 2006. "Before we can enact a national plan, Congress needs its own plan to properly dispose of its own e-waste," says Thompson. The bill comes at a time when manufacturers, recyclers, and consumers are debating the issue of establishing a national standard that would keep consumer electronics out of the junk pile. Having consumers pay for recycling in the form of an upfront fee has gained the support of Rep. Louise Slaughter (D-N.Y.) and the Consumer Electronics Association, which has a membership that includes electronics makers and sellers. However, Hewlett-Packard favors a take-back model that would give consumers an opportunity to return old products to manufacturers that are willing to invest in e-waste recycling infrastructure.
Click Here to View Full Article

U.S. students are laboring under the false assumption that IT jobs are scarce, dull, and antisocial, and this misperception must be corrected if the decline of U.S. computer science majors and the IT workforce is to be reversed, writes Maria Klawe, dean of the School of Engineering and Applied Science at Princeton University and former ACM President. She says that without such a reversal, the U.S. is in danger of losing its global innovation lead. Data from the U.S. Department of Labor shows that even more computing-related job opportunities are available now than during the previous peak five years ago; furthermore, IT plays a vital role in all functions of the business community, which means that IT jobs, contrary to popular belief, are stimulating and require people who possess abilities far beyond mere technical skill. Klawe explains that current high school and university teaching methods fail to engage students who desire to incorporate their interest in computer science with other skills and interests, and also exclude women and minorities for the most part. She recommends that an interest in math and science be nurtured in grades K-12 through students' exposure to real-life projects and careers in engineering and applied science. Klawe outlines a strategy for improving college-level computer science education that involves making such courses an undergraduate requirement; ensuring that students are excited by their first computing course; setting up links to other disciplines with double majors and multidisciplinary programs; and developing industrial alliances and internships so students can gain a more accurate view of IT careers. To encourage greater participation of underrepresented groups, the author urges the hiring of more diverse computer science faculty and the provision of much-needed role models. Klawe concludes that long-term support from industry and government is essential if such efforts are to be successfully implemented and sustained.
Click Here to View Full Article

DePaul University computer scientist Tom Muscarello has developed the Classification System for Serial Criminal Patterns (CSSCP), a computer system designed to help police investigate violent crimes, especially serial crimes. Due to go live as early as next year, CSSCP employs pattern-recognition software to aggregate crime data, such as the age, gender, and height of the perpetrator, and the types of weapons and vehicles used, to generate a criminal profile. The system aims to overcome the trouble humans have finding patterns to connect serial crimes, a principal shortcoming in detective work. CSSCP can sift through thousands of criminal reports each second, scanning for repeated phrases or words that would go unnoticed by human detectives. Serial crimes are particularly difficult to solve because they often happen over a long period of time and across a large area, sometimes involving police of different jurisdictions. The system was patterned on the techniques of six of Chicago's best detectives, said Muscarello, who has been working on artificial intelligence technology to help police since the mid 1990s. Built on the Kohenian neural network, the CSSCP system found at least 10 times the number of linked crimes as a group of detectives working with the same data in a recent trial. While other computer-based detective aids exist, "nobody today in 2005 has come up with a program that has done what this network can accomplish," said Charles Padgurski, formerly of the Chicago Police Department.
Click Here to View Full Article

Affective computing systems can read users' emotional states and adjust their operations to accommodate those states. One example is Laura, an "emotionally intelligent" virtual fitness trainer that encourages users to stick with an exercise program via a combination of friendly gestures and dialogue, empathetic facial expressions, and a soft, soothing voice. Her developer, former MIT graduate student Timothy Bickmore, explains that Laura's personality is programmed within an augmented transition network, which he describes as "a series of decision trees that represent different fragments of the conversation." The software assesses a user's answers to Laura's questions to determine the best response; analyzes the factual content of what the trainer will say and chooses expressions, hand gestures, vocal tones, and other emotional cues in which to render the response; studies the content of what it is about to say, compares it to previous dialogue, and then ascertains which elements of the dialogue contain new data; and physically emphasizes the speech fragments with new information. However, users must enter their dialogue through a keyboard instead of a voice interface, while another challenge Laura's developers are working on is the development of software that can read emotional signals indicating when intercession is--and is not--the best course of action. Among the critics of affective computing is virtual reality pioneer Jaron Lanier, who warns that investing computers with the guise of empathy could consequently make users "emotionally stupid."

The many triumphs of Central European computer programmers in international contests underscores the region's growing status as a hotbed of exceptional technical talent, whose appeal to knowledge-driven industries is increasing thanks in part to cheap labor costs and the increasing turnout of skilled professionals. Central Europe rakes in $37 billion in foreign investment yearly from the telecom, automotive, and pharmaceutical industries, to name a few; the area is also starting to gain a foothold in the engineering and software development sectors. IBM ranked Hungary, Poland, and the Czech Republic among the 10 leading global destinations for research and development jobs in the first six months of 2005, while multinationals such as Hewlett-Packard have set up or are setting up facilities or branches in Central European nations. Most remarkable is Central Europe's meteoric success after decades of mismanaged government, economic despair, and crime under Communist regimes and the destabilization that came with the collapse of the Soviet Union. One of Marxism's positive legacies to the region is outstanding universities and technical schools, which graduate workers trained to think creatively, an advantage cultivated in the days of Communism when resources were scarce. The prestige of these schools is one of the factors luring foreign investors to Central Europe, while lower labor costs than Asian countries is another. The admission of Central European countries to the European Union is also enhancing their attractiveness to investors. The most distinguishing feature of Central Europe is the ambitious and competitive streak of its youth.
Click Here to View Full Article

Congress is expected to pass privacy legislation in 2006 to fight the ongoing problem of identity theft, which the American government estimates costs as much as $50 billion a year. Laws aimed to protect privacy in America are gaining support from lawmakers and businesses alike. At the Congressional Internet Caucus, Microsoft general counsel Brad Smith said the company has changed its mind and now backs a robust national privacy law to apply to all companies, online or offline. Smith said current rules covering data protection are overlapping and inconsistent, so a federal law would make it easier for businesses and secure citizens' trust. Around 20 bills were drafted this year after a series of leaks at credit card data processing firms and other businesses exposed the personal data of more than 50 million consumers. A California law forced the incident to be disclosed, and is influencing businesses to put pressure on Congress to adopt a similar law nationwide. Previously, America was at war with the European Union over its privacy laws that require companies to elaborate on what data they are collecting and how it is controlled. The European Union is now considering adding its own requirement, similar to the California law, to strengthen its position as the leader in privacy legislation.

Pervasive computing in automobiles, a hybrid form/function rapid prototyping environment, recognition of situations and recurring contextual data, and high-level workflow paradigms for device interactions are areas of research in the building of pervasive computing prototypes. Students at the University of Buffalo's MediaRobotics Lab engaged in several experiments to research pervasive computing applications for road trips through the use of prototyping kits; in one project, a student equipped toy animals with equipment that recorded scenery changes, which were wirelessly relayed to a garage where returning passengers could view a free-form rendition of the data captured during the trip on a screen. Seng Loke of Monash University's Caulfield School of Information Technology is working on situation pattern language for recognizing scenarios that could help software developers address recurrent problems. Loke and fellow Caulfield student Sea Ling are also developing a high-level model to describe interactions between devices in accordance with workflow paradigms. Users can convert scripts written in the researchers' high-level language into a specialized Business Process Execution Language for Web Services specification, and then implement the scripts using a device ecology workflow engine. The d.tools integrated design environment being developed by students in Stanford University's HCI Group gives designers the ability to quickly and easily create combined form/function prototypes early in the design process. The environment comes with a library of plug-and-play hardware elements and a visual authoring space for developing interaction models.
Click Here to View Full Article