A team of engineering students in the ACM'sUCLA chapter is developing a virtual communication system for classmates using the open source Jabber instant messaging protocol. The CourseChat service is being created in response to the School of Engineering's request to embed instant messaging within the CourseWeb content management service for the school's class Web sites. The students utilized a modified version of the open source client Jbother to link to the Jabber server; Jbother is written in Java, and will work on any computer equipped with a downloadable Java virtual machine. UCLA engineering student and CourseChat co-developer Shaun Ahmadian says students could use the service to hold virtual study groups from their home computers, while professors or teaching assistants could use it to put in hours at a "virtual office." A student who accesses CourseWeb can log on to CourseChat and view a list of all students in their classes, and then initiate private as well as group discussions. Requiring students to log on to CourseWeb prevents them from making bogus or multiple aliases, and only allows them to access rooms for classes in which they are enrolled. The Jabber team sees its recently completed whiteboard function as an especially valuable tool for students: With whiteboard, they can draw diagrams and equations with their computers, and concurrently display their work on another student's screen. Ahmadian thinks many scientific disciplines will benefit from this tool by allowing students to better visualize concepts under discussion.
Click Here to View Full Article

World Wide Web Consortium (W3C) director Tim Berners-Lee told attendees at the Fourth Annual Bio-IT World Conference and Expo on May 17 that the Semantic Web could enable researchers and other users to uncover scientific data concealed within dissimilar applications' formats and organizational constraints. "At the moment, I see a huge amount of energy from people in life sciences getting excited by the Semantic Web and what it can do to solve the big-idea problems," he said. Berners-Lee believes the life sciences will push development of the Semantic Web. He said, "If we could get critical mass in life sciences, if we get a half a dozen or a dozen set of ontologies, the core ones for drug discovery out there, then suddenly the Semantic Web within life sciences would have a critical mass" that would inspire other disciplines to invest as well. Berners-Lee noted that drug researchers use databases and information systems that are already in machine-readable formats or are ready to be converted into such formats. However, Berners-Lee cautioned that those formatting their data to conform to the Semantic Web are not guaranteed a rapid return on investment. Representatives from the University of Colorado, the W3C, Oracle, IBM, and other organizations have developed Biodash, a prototype Semantic Web drug development dashboard that draws associations between diseases, drug progression stages, molecular biology, and pathway knowledge. The tool features a Semantic Web browser for linking data from public sources and chemical libraries with genes, proteins, pathways, and other biological entities.
Click Here to View Full Article

University researchers have developed a new method for efficiently accepting large amounts of incoming data by borrowing from genetic algorithms. University of Southern California computer science professor Leana Golubchik says government agencies face a serious technical problem when large amounts of data must be distributed among multiple machines, such as when the IRS has to deal with tax returns near the April 15 deadline. The new data transfer technology, called Bistro, draws lessons from genetic algorithms to ensure data reliability and the efficient use of resources: Under the Bistro framework, end users upload data to distributed computers, called Bistros, with each node generating a receipt that verifies the integrity of that set of data; these receipts are collected by the agency instead of the data itself, greatly reducing central server loads. Bistro specifically draws on genetic algorithms by fully utilizing the best computers and randomly mixing a small number of data fragments between pairs of computers, a process that mimics evolutionary mutation. The results were compared against traditional methods for distributed data storage systems, such as depending on select computers, spreading data evenly across all computers, and assigning data to computers based on their reliability. Although Bistro was not as accurate as the so-called brute-force technique, that method requires excessive processing resources, and Bistro is a good shortcut for achieving efficient results, says Golubchik.
Click Here to View Full Article

Shanghai Jiao Tong University's last-minute, first-place win of the ACM's International Collegiate Programming Contest in April demonstrates how computer-programming students in developing nations are acquiring superior skills thanks to the rapid proliferation of information technology. The competition involved 78 teams of college students working to solve 10 real-world programming problems in five hours, using the Eclipse operating system and programming languages such as C++, Pascal, and Java. The winner was the team that solved the most problems in the shortest amount of time and the least number of attempts. The coach of the winning team said the students devoted a great deal of their extracurricular time in preparation for the contest. In the competition's early years, contestants were strictly European and North American, and contest executive director William Poucher attributes this to the fact that in those days, "there were only a few hundred computer science professors in the world--and we all knew each other." Students from current and former communist countries made up nine of the 12 finishing teams in this year's finals, while teams from Canada, Sweden, and Norway comprised the remaining three. Poucher does not think this signifies a decline in Western students' programming prowess, because competition is so fierce at the finals level. More distressing is the underrepresentation of women in the contests, and director of IBM's worldwide university talent programs Margaret Ashida says women's disinterest must be investigated and counteracted.
Click Here to View Full Article

For more information about ACM's International Collegiate Programming Contest, visit http://icpc.baylor.edu/icpc/.

The Australian government has failed to increase the number of women in the IT field because it focused on barriers to recruitment instead of the qualities of the IT industry that are off-putting for women, says Australian Computer Society (ACS) Women board program director Su Spencer. The Australian government warned about low percentages of women in IT in a March 2000 report, but little progress has been made in the meantime; the Australian Bureau of Statistics says about 20 percent of the Australian IT workforce is female, and Spencer says the failure to increase the proportion of women in the workforce means it is time to re-think strategies and assumptions. The biggest needs are for women IT workers to set role models for others to follow, and for employers to craft new workforce policies that take family considerations into account. Technology can also play a role in creating a more hospitable work environment, such as by enabling telecommuting or opportunities in open source development, where contributors interact without regard to gender, race, or age. Although stereotypes of "geek isolation" remain an obstacle to more women entering the IT workforce, once people enter the industry they discover there are many avenues IT experts can take. Not every IT job deals with technical complexities: User-interface design, IT marketing, and business analysis are all outside the traditional stereotype of a software programmer.
Click Here to View Full Article

For information on ACM's Committee on Women in Computing, visit http://www.acm.org/women.

University at Buffalo computer scientist Matthew J. Beal's forte is Bayesian machine learning, a field of study that seeks to build models that computers can use to become more knowledgeable through probability inference. Beal, who joined the Department of Computer Science and Engineering in UB's School of Engineering and Applied Sciences last fall, devises machine-learning algorithms that could be employed in a diverse number of fields, including artificial intelligence, bioinformatics, document-information retrieval, pattern recognition, and human-computer interaction. His algorithms could one day help scientists monitor the time evolution of specific cells to aid in cancer research and other kinds of research, for example. During a brief stint at Microsoft Research, Beal designed and constructed Bayes Nets for multimedia integration and administered graphical model inference algorithms and Bayesian learning to the concurrent fusion of audio and video data for optimal-tracking operations, and also created Bayesian hierarchical wavelet state-space models for video sequence compression. His current research and documentation emphasizes such subjects as variational Bayesian techniques, hierarchical Dirichlet processes, embedded hidden Markov models, microarray analysis with variational Bayes, a tool for time series inference, and probabilistic sensor fusion. Beal is a member of the Data Intensive Analytical Bioinformatics Core Group of UB's New York State Center of Excellence in Bioinformatics and Life Sciences, and was a program committee member for the International Conference on Machine Learning in 2004. This year he will be a program committee member for Uncertainty in Artificial Intelligence and Artificial Intelligence and Statistics.
Click Here to View Full Article

UCLA engineering professor Rajit Gadh is developing a DVD-protection scheme that taps the ability to write data to radio frequency identification (RFID) tags. Upon the purchase of a DVD, the buyer would be required to provide a password or some kind of biometric data that would be written to the disc's RFID tag. The consumer would then have to re-enter this information into a special DVD player in order to play the disc. Gadh says there is no question that such a system would be more restrictive to DVD users than anti-copying measures such as the Content-Scrambling System, which has already been hacked. Princeton University computer science professor Ed Felten doubts that customers would be willing to accept such a scheme, arguing that "people would find it creepy to give their fingerprint every time they wanted to play a DVD." He also sees little sense in purchasing an RFID reader-equipped player in order to buy less functional DVDs. Meanwhile, privacy proponents are worried that RFID tags can connect products to individuals without their knowledge. Electronic Frontier Foundation staff technologist Seth Schoen is doubtful that Gadh's digital rights management scheme will be any more effective than its predecessors, noting that "It only requires one person to break it."
Click Here to View Full Article

The IST-funded FAME information butler is an intelligent agent designed to help groups of people overcome linguistic, cultural, communication, and information challenges while collaborating on common tasks. "The information butler facilitates human to human interaction, eliminating the need for people to interact directly with a computer to obtain information and allowing them to work without necessarily being conscious that a machine is working alongside them," says project coordinator Florian Metze of Germany's University of Karlsruhe computer science department. The butler picks up keywords from users' conversations by microphone and processes them with a distant speech recognition system. The agent makes connections between the keywords and information in the system database or data uploaded from the Internet, and presents this information to users. Users can interact in potentially any dialect through speech recognition and real-time translation technology, which also enables the automatic transcription of conversations for future citation. The demonstrator consisted of a wall display and a touch-screen Augmented Table that displayed tokens related to different information and topics being discussed.
Click Here to View Full Article

The Defense Advanced Research Projects Agency's (DARPA) Grand Challenge is a race between autonomous vehicles that must navigate a 175-mile Mojave Desert course without human assistance; the team that develops the first vehicle to cross the finish line in less than 10 hours will win $2 million. This year's entry from Stanford University is Stanley, a modified Volkswagen Touareg. Members of the Stanford Racing Team are testing Stanley in the desert terrain of Barstow, Calif., and they plan to log 2,000 autonomous miles in preparation for the race. The specific route of the race will not be disclosed until just two hours before the start. "This is the first endurance race in history where the machine will have to make all the decisions," says team leader Sebastian Thrun, associate professor of computer science. Stanley can calculate its position within two inches via sensors and global positioning systems, while lasers on the roof scan the ground ahead for obstacles and in-vehicle cameras take snapshots of the course; when traveling at racing speeds up to 35 mph, the vehicle scans the horizon with long-range radar, and it "learns" the terrain and maps out an optimal driving surface with computer algorithms. Stanley evaluates the relative safety of multiple prospective routes and deciphers sensor input through a probabilistic algorithm. Seven ruggedized Pentium M laptops in the trunk provide the vehicle's brainpower, and the machines are wired to an electronic brake, throttle, and steering column; doctoral candidate in computer science David Stavens says this setup ensures that Stanley will be able to continue in the race if one or several computers malfunction. DARPA launched the Grand Challenge to accelerate the development of robotic vehicles as part of its effort to automate one-third of all military vehicles by 2015.
Click Here to View Full Article

COBOL is not necessarily a thing of the past, considering that use of the mainframe has been growing and the enormous amount of the COBOL code currently running. In fact, programmers who are skilled in the legacy language are likely to be in high demand over the next decade when many COBOL programmers retire. Preliminary results of a Micro Focus International survey of 750 mainframers in the United States and Canada indicate that the median age of the COBOL programmer is 45-49. When the final results of the survey are made available in June they are expected to show that 52 percent of mainframe applications are still written in COBOL, and that 41 percent of mainframers consider COBOL to be a principal programming language by a margin of about 25 percent over Java. Ron Kizior, a Loyola University Chicago School of Business professor who was involved in the research, says the skill that will be in demand in the years to come will be the ability to integrate COBOL with Web-oriented development tools. Only 10 percent of mainframe shops were found to have gone to the Web. "Some of the more enlightened universities are trying to get to grips with this emerging requirement, and are bringing the necessary mainframe components back onto the syllabus," says analyst Mark Lillycrop.
Click Here to View Full Article

Experts expect new video search engines to significantly enhance corporate intelligence gathering and help spawn new applications and products such as phonetic-based Web searches and Internet-specific entertainment from Hollywood, among other things. Clew partner David Carpe is particularly enthused about how corporations can employ video search to compile intelligence on their rivals, noting that "major brands are interested in monitoring mentions of their own products and those of competitors." The current corporate intelligence acquisition option--media monitoring services--is costly, and improved video search engines will allow enterprises to establish their own search parameters. Carpe believes hybrid video-audio search services will be advantageous for consumers and other users, and says such services will make it easier for people to locate procedures, instructions, or demonstrations. Broadband users stand to benefit from Internet sites' efforts to enable their content for video search; CNN.com, for instance, plans to offer a video browser and search options that will let users locate and organize videos, concentrating on the content they wish to view. Meanwhile, MeeVee is working on a personalized TV search technology that allows consumers to find shows they want to see, to the exclusion of undesirable programming. A MeeVee representative says users would be able to search for programs based on keywords, show titles, character names, or cast members, and even add preferred shows to their personal planners so they can watch them. A search engine from Nexidia can carry out phonetic-based searches of audio content, removing the need to spell items properly.
Click Here to View Full Article

The Transaction Processing Performance Council is set to replace a number of server benchmarking tests, including the 13-year-old TPC-C test that measures large database servers. TPC-C will be replaced by TPC-E in 2006, enabling prospective server buyers to more accurately gauge the real-world performance of machines; the new test will also be less expensive to run and less easy to optimize configurations for. TPC-C has long been the most widely referenced statistic in request-for-proposals, and is preferred over other tests because it provides price to performance ratios. In the past, server vendors have gone to great lengths to improve their TPC-C scores. TPC-C is meant to simulate a warehouse inventory database and counts transactions per minute. Those numbers have increased tremendously in the past decade, first reaching about 100,000 in 1998, then growing to 3.2 million transactions per minute with an IBM server in 2004--but that machine utilized a stupendous 6,548 hard drives, demonstrating the increasingly esoteric quality of TPC-C testing. The new TPC-E test will instead simulate an electronic brokerage that performs more sophisticated processes and is more realistic in that it is harder to run on distributed computing clusters. The council also plans to release a new TPC-App test for application servers, which will help compare .Net and Java performance, and a TPC-DS data warehouse test.
Click Here to View Full Article

Upwards of 100,000 open source projects have been collaboratively developed and stored on SourceForge.net, whose registered users numbered a record 1,074,424 as of May 15, 2005. Among the companies and organizations posting successful projects to SourceForge.net is NASA, whose Java Pathfinder program helps pinpoint bugs in other programs by running test trials until a program failure is induced; Google, whose contributions include Google mAIM, Sparse Hashtable, GoopyFunctional, CoreDumper, and Perftools; SalesForce.com, which has posted samples and utilities that can be used with the sforce Web services platform; and IBM, which has initiated new online skills-building programs to drive innovation, collaboration, and development around nascent open source projects. "The community that relies on our site represents the earliest of early-adopters--the most forward-thinking development and IT professionals who are seeking better, faster, and frankly, cheaper, ways to achieve their business goals," says OSTG VP and SourceForge.net site director Jeff Bates. VA Software CTO Colin Bodell says, "What is especially exciting is that we're seeing more and more software companies adopting open source business models and practices--this is not just a trend; it's becoming the way that all software will be developed, distributed and maintained." Large enterprises have publicly committed to open source as a vehicle for encouraging creativity and purchase toward next-generation technologies and services. A recent Forrester Research study finds open source software being used or planning to be used in 60 percent of firms interviewed.
Click Here to View Full Article

Improving the overall cybersecurity situation means preparing for failure by strengthening internal security and business continuity planning, not just focusing on perimeter security. Former White House advisor Richard Clarke says software vulnerabilities are the main threat to cybersecurity today and that government, academic, and industry experts should draw up new software development standards, whose existence would make software security a more marketable aspect. And while many large software vendors have already established much better coding practices, those processes are proprietary and cannot be independently tested by auditors. Even if perfect code was possible, it still would not solve the problem created by integration and fast-changing business requirements, says SystemExperts' Brad Johnson. Apart from coordinating better software development standards, Clarke says the government plays an important informational role and should consolidate and quickly disseminate data concerning attack methods, though many corporate IT leaders currently feel the government is a difficult partner in this area. CERT Chairman and eBay CISO Howard Schmidt says the government has already laid down good plans in terms of cybersecurity, but the difficult task of implementation needs to be tackled by the private sector. Another area of concern for Clarke is businesses' focus on perimeter IT security, which leaves their inner systems vulnerable; organizations need to prepare to be hacked, and can minimize disruptions by compartmentalizing systems, using intrusion detection and prevention at the desktop level, and strengthening identity-based access controls and user authentication. Clarke also says his experience during the Sept. 11 attacks taught him that government agencies should have alternate sites "warm," or at least partially staffed and ready to go in an emergency.
Click Here to View Full Article

Portland State University attempts to stimulate an interest in math, engineering, science, and computer science among underrepresented high school students through the 35-year-old Math, Engineering, and Science Achievement (MESA) program. Oregon MESA director Carla Faini says the program arranges weekly conferences between participating students and advisors, and hosts science fairs, SAT prep classes, industry site visits, and other special events and activities. May 13 was MESA Day at PSU, where visiting student teams participated in various challenges: One challenge was to design a powered vehicle from furnished materials in 50 minutes; another was to construct cars driven by a mousetrap. Still another was to build a bridge while adhering to strict design, structure, and size requirements, after which it was subjected to stress to evaluate its stability. Advisor to the Jefferson High School team Daniel Motta praised the MESA program. "I watch the kids building something, testing it, trying it, trying it again," he said. "They learn a lot that they'll end up using in life." At the May 13 event, assistant dean of PSU's Maseeh College of Engineering and Computer Science Marcia Fischer announced a scholarship in which any Portland-area MESA participant who enrolls in an engineering or computer science major at PSU will receive an annual stipend of $1,000.
Click Here to View Full Article

The U.S. Circuit Court in Washington recently determined that the FCC exceeded its bounds by trying to require TVs or other devices to read "broadcast flag" coding designed to block the copying of digital TV signals, but Chris Nolan writes that the copyright controversy will undoubtedly continue. Public Knowledge CEO Gigi Sohn sees a strong link between the broadcast flag case and the case of MGM v. Grokster currently before the Supreme Court: "It's about who's going to control technology," she says. The Supreme Court's ultimate ruling would impact not just peer-to-peer file sharing technology, but video and other media as well. The general feeling is that the Grokster case will be re-submitted to California's 9th Circuit Court, which Nolan expects will weigh the broadcast flag ruling in its decision. Congressional supporters of the television, recording, and film industries may lobby for an expansion of the FCC's powers by citing the broadcast flag ruling, while tech companies might use the as-yet-unannounced Grokster ruling as an argument that a line must be drawn concerning Capitol Hill's authority in shielding copyright owners from theft. Nolan says copyright legislation will take center stage in Congress following the Grokster ruling, but doubts the debate will end with the passage of a law. "What the industries don't get in copyright, they'll try to get in telecom: Two apples, many, many bites," she concludes.
Click Here to View Full Article

For information on ACM's activities regarding MGM v. Grokster, visit http://www.acm.org/usacm.

An organization can benefit enormously from end user software development, which is becoming increasingly necessary as untrained employees grow impatient with IT for the tools they require to access and manipulate data. National Council on Compensation Insurance CIO Shawn O'Rourke calls not just for the sanctioning, but also the encouragement, of end user development; he further recommends that IT's rejection of employees' requests for more tools should be accompanied by both an explanation and suggestions for suitable alternatives whenever feasible. Simplicity is also key in end user development, and minimizing errors requires an intuitive user interface such as "natural-language" programming based on familiar syntax or drag-and-drop and pull-down menu techniques. O'Rourke points out that "It takes a lot of interaction back and forth between IT and the business to see that [end user development tools] are deployed effectively." Isis Pharmaceuticals' John McNeil says most software projects fail in the requirements phase, when a gap between end users' needs and IT's ultimate capabilities appears. End user development can therefore lessen the hardship for IT, even if only a prototype rather than a working application is produced. However, the simplicity offered by end user development tools could give rise to unrealistic expectations for users who may want to prototype a complete system not because it is technically possible, but because it can be done. Consequently, it is a sensible strategy to let back-end connections remain the province of IT experts, while leaving interfaces and navigation to end users.
Click Here to View Full Article

In "Flight of the Creative Class: The New Global Competition for Talent," George Mason University public policy professor Richard Florida writes that the global economy's increasing reliance on knowledge industries requires the cultivation of a creative class by nation-states; he warns that the United States could lose its competitive edge because it no longer encourages the growth of such a class. Florida argues that "as a community, we need to start paying more serious attention to the creative ideas of our youngest, oldest and most marginalized members," adding that substantially higher investments in research and development and universities are necessary if countries are to attract the most talented people. Moreover, both the private and public sectors need to channel more capital into all creative and innovative disciplines, including art and culture. Florida also calls for a ramping-up of efforts to increase the market for creative opportunities by producing high-end jobs in R&D, arts and culture, higher education, and innovation. The author says the wintry economic climate is spurring state governments to scale back budgets for higher education, and the resulting reduction of opportunities undermines universities' image as models of tolerance and diversity; this in turn discourages foreign talent from enrolling in such schools or prompts them to return to their native countries once they graduate, and also encourages domestic talent to look elsewhere for educational opportunities. The overall result is a lack of creative talent desperately needed to reinvigorate communities, economies, and culture. "Only when we begin to see all of these investments--scientific, economic, artistic, cultural and others--as mutually reinforcing parts of the same creative whole will we begin to take advantage of even a fraction of our latent human potential," Florida contends.
Click Here to View Full Article

Many librarians and archivists have welcomed Google's announcement that it will digitize millions of library books, although some are worried that such an venture could lead to the privatization of literary knowledge. There is another concern that digital libraries will drive traditional libraries and librarians into obsolescence, but most library digitization advocates believe traditional libraries will benefit. UCLA librarian Gary Strong says libraries and industry are collaborators in the digitization effort, but maintaining the usability of information is a priority. Google product manager and Google Print project leader Susan Wojcicki believes digitized libraries could revolutionize activities such as linking related text, but she acknowledges that the Google project depends on in-development robot cameras that will be able to convert printed books into searchable Web pages with assembly line efficiency. Two copies of the digital book will be retained: One for Google and one for the partner library, which will be permitted to use the copy as it sees fit, provided it does not share the copy with Google's competitors. This process follows a model outlined by Internet Archive founder Brewster Kahle in which parallel private and public databases coexist, although Kahle is concerned that libraries will be prevented from working with other organizations or companies to distribute digital texts. Some librarians believe the definition of fair use will need to be rethought if digitization becomes widespread. There is also a feeling that digital libraries will keep librarians busy--perhaps too busy, given the problems with cataloguing and preserving digital assets, as well as assessing a much larger volume of content.
Click Here to View Full Article