Read the TechNews Online at: http://technews.acm.org
ACM TechNews
November 22, 2006

MemberNet
The ACM Professional Development Centre
Unsubscribe

Welcome to the November 22, 2006 edition of ACM TechNews, providing timely information for IT professionals three times a week.

Sponsored by Information, Inc.

http://www.infoinc.com/sponsorthenews/contactus.html


HEADLINES AT A GLANCE:

 

Programs Written in Old Code Pose Business Problem
Financial Times Digital Business (11/22/06) P. 6; MacKenzie, Kate

While legacy languages such as Cobol are still in relatively heavy use by many companies, the number of programmers able to work with them is dwindling. Between 12 and 15 percent of new development, mostly back-end financial systems, is being written in these languages, says Gartner's Jim Duggan, and a Computerworld survey showed that 58 percent of respondents using Cobol are developing new applications in it. Gartner estimates that 50 percent of programmers who are skilled with mainframes, on which legacy programs are run, will be eligible to claim their pension by 2007, and to make matters worse, these older employees also expect higher salaries. Many schools no longer teach legacy languages, favoring "object-oriented" languages. Many companies are turning to internal training of recent graduates, while some have even worked with nearby schools to develop legacy language education curriculums. Some are choosing to keep applications in legacy languages but run them using SOAs that are easier to maintain; "clone" products provide the additional option of reproducing legacy versions of software from mainframes onto modern systems. The fact that these languages have hung around shows the longevity of the mainframe, which many thought was coming to an end in the 1990s; mainframe giant IBM has established a new educational programs to promote and simplify mainframe administration.
Click Here to View Full Article - Web Link to Publication Homepage
to the top


Kurzweil: Computers Will Enable People to Live Forever
InformationWeek (11/21/06) Gaudin, Sharon

In a keynote speech at SC06, "The Singularity is Near" author, futurist Ray Kurzweil, explained his views that computer, or non-biological intelligence, will allow humans to overcome illness and aging in just 25 years. By reverse engineering our own brains, scientists will be able to develop intelligent machines that will surpass human intelligence. He describes the medical revolution he predicts as replacing the "human body version 1.0" through technology such as nanobots that can swim through our blood stream to make necessary repairs to keep us young, and computers that can download and backup our memory. "$1,000 worth of computation in the 2020s will be 1,000 times more powerful than the human brain," Kurzweil predicts. Computers will be integrated into the world around us, even our bodies, he adds; imagine a computer in your eye that projects images upon your retina. "We won't experience 100 years of technological advance in the 21st century; we will witness on the order of 20,000 years of progress, or about 1,000 times greater than what was accomplished in the 20th century," he said. Kurzweil has received the National Medal of Technology and the Lemelson-MIT prize; his book has been endorsed by Bill Gates himself, a robotic director at CMU, and a professor and a physicist at MIT. Kurzweil says, "Supercomputing is behind the progress in all of these areas," and predicts that non-biological intelligence created in 2045 will be one billion times more powerful than all human intelligence today. He also predicts that in 15 years virtual reality environments will be routinely used to talk with others instead of making a cell call and computers will be ubiquitous and pervasive.
Click Here to View Full Article
to the top


Electronic Voting Trend May Be Short-Circuiting
Sarasota Herald-Tribune (FL) (11/19/06) Hull, Victor

An audit of the congressional election in Sarasota County, Fla., is at the center of a push to require electronic voting machines to produce paper records, or to even ditch electronic voting altogether. Support is rising in Congress for legislation requiring a paper trail, and a bill has even been filed that would require a hand count for the presidential election. Twenty-seven states have already passed a paper-trail mandate, some also requiring audits of the electronic voting process. The Sarasota election was mentioned by Democrats in Congress and is seen by many citizens as a clear indictment of e-voting, since there is really no way of figuring out what went wrong, as e-voting expert Avi Rubin of Johns Hopkins University points out. Votersunite.org executive director John Gideon feels that an examination of the Sarasota problem will serve as "a death knell" for the technology. Verifiedvoting.org's David Dill believes optical-scan voting would have prevented the Sarasota problem, while others would only feel comfortable with hand-counted paper ballots; but both of these systems have been found to have their share of flaws as well. Officials such as Charlotte County, Fla., elections supervisor Mac Horton, whose district uses the same machines as Sarasota County, are reluctant to abandon the costly system. "I've been very well pleased," says Horton. If it's left up to me, I'd stay right where I'm at." For information about ACM's e-voting activities, visit http://www.acm.org/usacm
Click Here to View Full Article
to the top


Science Ph.D.'s Continue to Grow
Inside Higher Ed (11/20/06) Lederman, Doug

A new National Science Foundation report found that a record 27,974 science and engineering Ph.D.'s were handed out by American universities in 2005, eclipsing 1998's previous all-time high of 27,232, but this news will not be enough to stem the growing fear of weakening American scientific competitiveness. The number of Ph.D.'s given to women, Asian Americans, several underrepresented minority groups, and "STEM" fields were also found to be at an all-time high. However, the greatest increase in growth was in the category of non-U.S. citizens, who earned 13.4 percent more doctorates from American universities in 2005 than in 2004. Moreover, 41 percent of all doctorate recipients in 2005 were foreign-born, up from 39 percent in 2004. More Ph.D.'s were handed out in the U.S. in 2005 in computer science than any previous year, up significantly from 2002. Females accounted for 19.8 percent of computer science doctorates awarded in 2005, up from 15.1 percent in 1996.
Click Here to View Full Article
to the top


Image Labeling for Blind Helps Machines 'Think'
Washington Post (11/21/06) P. A2; Goldfarb, Zachary A.

An online game has been designed to make image labeling fun and make surfing the Internet easier for the blind. The type of Internet program used by the blind reads Web pages aloud, but since images cannot be identified by the program and thus have no way of being spoken, many pages are prohibitive. The solution to this is image labeling, which would give these programs a way to describe images verbally. Carnegie Mellon University computer science professor Luis von Ahn developed the ESP game for just this purpose: random visitors to ESPgame.org are paired up and challenged to provide identical labels for the image they see; some people have spent as much as 40 hours a week on the site. Programs such as the ESP Game are known as human computation, where a computer asks a human a question and the human does the answering. Teaching a computer using human computation is a lot like the way children learn to identify things, but as von Ahn says, "Nobody bothers to teach a computer." CMU's Manuel Blum, who advised von Ahn's dissertation, explains, "What he's doing is mining the ability of humans." Von Ahn aims to develop computer intelligence that resembles that of humans and could perform language translation that accounts for the subtleties of foreign languages, for example, or make fast illness diagnoses in hospitals. Von Ahn says his goal is "To be able to use all of this data and to have computers be able to do pretty much everything we can do."
Click Here to View Full Article
to the top


Networks Could Self-Organize Sooner Than We Think
IST Results (11/21/06)

Autonomic communication, which is expected to help networks automatically adapt to the growing complexity of the Internet, has received a boost in recent years from technology known as "self-organization." The independent Automated Communication Forum (ACF), which was put together by European scientists in 2004, aims to enhance this new form of self-organizing, self-managing, context-aware, and autonomous type of networking through work in both market-related and non-market-related initiatives. Automated communication encompasses these "self" terms, as well as many other concepts, and according to Mikhail Smirnov, who has played a large role in the ACF and coordinated the ACCA project, a Future and Emerging Technologies initiative that came to an end this September, these technologies can ensure that networks have online identity, allow ISPs to better manage wholesale interfaces, and remedy the big problem of complexity in future networking technology by providing management. Self-management, says Smirnov, is the "top of the pyramid of technologies...It's not about putting intelligence into hardware, but making networks behave intelligently without human intervention." While these communications are currently accessible through an ISP or gateway, "in [the] future people will be connected directly to their contacts," which will require enhanced linkage between users, content, and service on the part of telecoms, explains Smirnov.
Click Here to View Full Article
to the top


Phishing Toolbars: All as Hopeless as One Another
Techworld (11/20/06) Dunn, John E.

Anti-phishing Web browser toolbars are not very effective, concludes a new study conducted by Carnegie Mellon University researchers and supported by the National Science Foundation and the U.S. Army Research Office. The study, "Finding Phish: An Evaluation of Anti-Phishing Toolbars," looked at 10 browser toolbars to determine their anti-phishing abilities and concluded that even the most capable toolbars (Earthlink, Google, Cloudmark, MS Internet Explorer 7, and Netcraft) identified only 85 percent of malicious Web sites, while the rest of the toolbars (eBay, Geotrust's TrustWatch, Stanford University's Spoofguard, and McAfee's Site Advisor) scored below the 50 percent mark. "Overall, we found that the anti-phishing toolbars that were examined in this study left a lot to be desired," said the authors of the study. "Many of the toolbars tested were vulnerable to some simple exploits as well." A good deal of those tested delivered a significant amount of false positives, which the researchers viewed as equally harmful because of the lack of trust this could breed in users. The researchers concluded that all filters must be used with care, and that the filter itself, not the browser it is used with, determines the level of security; the ability of the heuristics applied to detect fraudulent sites, and the usability of the software design for the user are the most important aspects of security.
Click Here to View Full Article
to the top


Hard-working Chips May Reveal Encryption Keys
New Scientist (11/20/06) Knight, Will

"Branch prediction" could put the modern microchip at risk to hackers, according to Jean-Pierre Seifert of the University of Haifa in Israel and the University of Innsbruck in Austria, and colleagues. Microchips second guess the logical flow of a program before the actual execution from branch to branch as a way to process information at a faster rate. However, branch prediction can tip off hackers about encryption key details that are processed, if there is a rapid increase in the work it performs and the time required, which would result from a need to perform another operation or a mistake. In a few thousandths of a second, Seifert and his team were able to figure out a high-security 512-bit encryption key, which is often used to protect online financial information and email messages from eavesdroppers. "Security has been sacrificed for the benefit of performance," says Seifert, who suggests the "Simple Branch Prediction Analysis" attack method could be carried out by hiding a small piece of software on a target computer. The researchers have posted their work online, and will participate in the RSA Security conference in February 2007.
Click Here to View Full Article
to the top


CMU Robot Car to Face Urban Traffic Challenges
Pittsburgh Post-Gazette (11/20/06) Templeton, David

The CMU Tartan Racing Team has one more year to perfect its robotic Chevy Tahoe for the 60-mile DARPA Urban Challenge, to be held next November at a yet-to-be-named Western location. Work is taking place at CMU's state-of-the-art Robot City, a converted steel site, where the car has already logged over 100 miles, including 50 "blind" miles. Although Congress will not provide any prize money for the winning team, DARPA's Track A teams, including CMU, can qualify for up to $1 million in funding. The competition asks each team to design a robotic vehicle that can navigate its way through a city, obeying traffic laws and dealing with traffic and other obstacles; "It's a whole other layer of complexity," said Tartan Racing's director of technology Chris Urmson. "People are working on perception problems with city speeds and urban driving conditions. Not a lot of work has been done in operating in those spaces." The vehicle must even be able to tell when a broken down competitor is in its way, realizing that it is then acceptable to break the double yellow line. Technology being developed includes 360-degree vision and depth perception as well as software that lets the car comprehend and react to what it sees. CMU has 24 researchers working on the project, and Tartan Racing director Charles "Red" Whittaker says, "Nothing less than full commitment will succeed. We're involved in it because it matters to the future of robotics, it matters to each of us individually, and we're in it for the win." Whittaker says to meet the demands of the race it will require the development of technology that is "a leap beyond what's already been done," but the ultimate payoff could be the creation of vehicles that do all or most of the driving and much safer driving conditions.
Click Here to View Full Article
to the top


Driving a Wheelchair With Your Shirt
Technology Review (11/20/06) Singer, Emily

Scientists at Northwestern University are creating a garment with built in sensors that can adapt to the way of quadriplegics, rather than previous systems that required patients to "fit the capacity of the machines," says Alon Fishbach, who works on the project. Control mechanisms currently used by quadriplegics include the sip/puff switch, which only allows two commands, and a headswitch that registers head movements against the back of the chair, but the Northwestern team's system is different. The shirt contains 52 flexible, piezoresistive sensors, developed at the University of Pisa, that change voltage as a result of being stretched at different angles. An algorithm has been developed to analyze the signals from each sensor in order to determine a definite number of movements that are translated into movement of the wheelchair. A virtual reality environment allows the patient to orient himself with the controls and also suggests ways to control the wheelchair more efficiently. The technology has been successfully tested with one patient, who has use of his hands. Further innovations are planned to bring this liberating technology to other aspects of life, such as video gaming.
Click Here to View Full Article
to the top


Quest for the Last Word in Search
Times Online (UK) (11/19/06) Durman, Paul

Google's director of research Peter Norvig describes today's search engines as "the very beginning of search," seeing the eventual possibility for semantic search engines. Google is constantly tweaking its search engine, not only combating companies trying to boost their own search rankings, but improving the effectiveness of the search engine's ability to provide users with appropriate, helpful results. Norvig, an artificial intelligence expert who previously worked at NASA, says that most of his colleagues are "disappointed" that searching still consists of entering a few words into a box. Google's goal is to be able search all types of information, including books, pictures, and videos. The company's Google Print project, for example, is an effort to digitize and make searchable thousands of books, while it's also working on speech-recognition technology that would allow a user to simply "tell" their cell phone what they are looking for and have Google provide them with results. Norvig says, "We need to understand where words are on a page, we need to understand diagrams. We need to update our algorithms to understand what the right answer is from a 500-page book. We are doing all that." One day Norvig hopes to see the advent of semantic searches, where users can simply pose a question and receive an answer. Meanwhile, other companies are attempting to surpass Google's abilities, including Hakia, a New York Firm advised by Yorick Wilks of Sheffield University, which is creating what they call a "meaning-based" search engine. Norvig believes that competition will only strengthen overall innovation, and that only Google is fully capable of realizing and deploying such innovations on a large scale.
Click Here to View Full Article
to the top


Wanted: Solutions for Post-CMOS Era
EE Times (11/20/06) LaPedus, Mark

Researchers at the AVS International Symposium & Exhibition in San Francisco expressed the need for a feasible replacement for CMOS-silicon within the next decade, when Moore's Law is expected to come to an end. Carbon nanotubes, nanowires, molecular electronics, quantum computing, three dimensional transistor designs, and spintronics are all possibilities. No format has received an overwhelming share of funding, but carbon nanotubes are being talked about the most; and a Massachusetts company has built and tested a 22-nm nonvolatile random access memory switch based on a carbon nanotube matrix structure set across an etched trench. However, Intel says that although nanotubes can be developed and utilized in a lab, mass production is still quite a challenge. Others favor spin technology, and the Nanoelectronics Research Initiative plans to develop devices with critical dimensions under 10 nm; but manufacturing of this technology has its share of difficulties as well. Self assembly is one possibility, as Wisconsin University scientists discovered a material called block polymers that can spontaneously assemble into complex 3D shapes when inserted onto special 2D surfaces. The future of the chip industry is uncertain, as 450mm fabs and 675mm fabs have been hypothesized, but would mean incredible costs.
Click Here to View Full Article
to the top


Sending Touch, Smell Over Net
Nikkei Weekly (11/13/06) Vol. 44, No. 2260, P. 16; Matsuda, Shogo

Researchers have set their eyes on the sense of touch, smell, and taste as they attempt to usher in a world of total perceptual communications. Sensual interaction over the Internet is primarily limited to seeing and hearing, but researchers across Japan are working to allow more sensory feedback for users of communications technology. For example, a tactile mechanism would enable surgical robots to transmit tactile sensations to a remote surgeon who performs a procedure while controlling surgical instruments on a video monitor. A pair of forceps, a scissor-like device developed by researchers at Keio University, is designed to allow the surgeon to "feel" the tissue and organs she is touching. "This promises a huge jump in the safety of robotic surgery," says professor Kohei Onishi of Keio. Meanwhile, the Tsuji Academy cooking-school chain has developed a device that can produce artificial smells like beef stew and curry using a process that is similar to the way in which printer ink is released from a cartridge. And Intelligent Sensor Technology is already bringing to market a taste sensing system that uses technology developed by Kyushu University's Kiyoshi Toko.
Click Here to View Full Article - Web Link to Publication Homepage
to the top


Grid and Bear it
Government Computer News (11/20/06) Vol. 25, No. 33, Jackson, Joab

Although grid computing has been widely adopted by the research community, it still struggles to find its place in the world of business. While grid computing is still young, even those who use it admit how complicated it is to effectively deploy, with problems ranging from high costs to a lack of applications. The technology has always been rather esoteric, beyond the reach of most IT shops, and as Ian Foster, who originally created the idea of grid computing along with Carl Kesselman, admits, "There is no turnkey solution, and there won't be one for a while." The Defense Information Systems Agency (DISA) has chosen on-demand computing, rather than grid computing, to meet its needs. The agency awarded contracts, that could be worth as much as $700 for the development of on-demand processing capabilities. Despite DISA's rejection of grid computing, the technology is gaining popularity in the pharmaceutical industry, and oil companies and financial firms also are using the technology to some degree. A major step toward establishing a presence in private business is for grid computing to cultivate partnerships with Web services. Kesselman is confident that in-roads will be made as a wider range of groups realize the potential benefits of grid computing. He says, "I think a lot of the early perception was that grid computing was for doing science and supercomputing. We've just begun to scratch the surface of real business transformation."
Click Here to View Full Article
to the top


Categorizing Web Search Results Into Meaningful and Stable Categories Using Fast-Feature Techniques
ResourceShelf (11/21/06) Kules, Bill; Kustanowitz, Jack; Shneiderman, Ben

Bill Kules, Jack Kustanowitz, and Ben Shneiderman of the University of Maryland's Human-Computer Interaction Lab and Department of Computer Science propose a number of "fast-feature" methods to categorize Web search results into stable and meaningful categories. These techniques were developed to address the metadata challenge of increasing numbers of unstructured and semi-structured digital documents, and the advantages such techniques yield include the provision of overviews, navigation within search results, and negative results. These methods use nothing beyond the features available in the search result list (title, snippet, URL, etc.), while credible knowledge resources (the Open Directory Project Web directory's thematic hierarchy, a U.S. government organizational hierarchy, personal browsing histories, DNS domain, and document size) are also employed to augment search results with important metadata. The researchers ran three tests in which the percentage of results categorized for a quintet of representative queries was high enough to suggest that the techniques were practically beneficial for such applications as general Web search, government Web search, and the Web site of the Bureau of Labor Statistics. A prototype search engine (SERVICE) incorporates fast-feature techniques, and Kules et. al make suggestions about improving categorization rates and how Web site designers could restructure their sites to support rapid search result categorization. They note, for example, that categorization engines would be capable of classifying pages in precise accordance with the authors' intentions if sites published a machine-readable site map and placed it in a standard location.
Click Here to View Full Article
to the top


Semiometrics: Applying Ontologies Across Large-Scale Digital Libraries
University of Southampton (ECS) (11/10/06) McRae-Spencer, Duncan M.; Shadbolt, Nigel R.

Duncan McRae-Spencer and Nigel Shadbolt of the University of Southampton's School of Electronics and Computer Science stress the need for services that can integrate and carry out inference calculations on the metadata generated by increasingly populous, accessible, and complete online digital libraries, and note that relational database management systems (RDBMS) cannot always deal with concurrent data updates and retrieval at immense scales. They think an alternative can be found in the expansion of RDF and the growing interest in Semantic Web technologies, and propose a method for large-scale metadata analysis and scalability testing that employs real-world data. The authors detail how RDF data storage and SPARQL querying offer performance levels at least as practical as standard SQL tactics for complex queries; they also have the flexibility and speed required for online digital library services. McRae-Spencer and Shadbolt conducted tests that determined that most searches translated into SPARQL were finished in a suitable time for use in Web services, while those that were too slow were aligned with the few that SQL queries could respond to in a reasonable timeframe from the open database. Blending the success of the SPARQL semantic Web service querying model with the SQL queries to the open database yielded the development of a Semiometric viewer application. "The results...show that in practice, the only realistic way for the SemioViewer application to work is to have both open SQL and SPARQL queries," the authors note. "While not typical Semantic Web applications, both the SemioViewer and the SPARQL-based Web services and client pages...require both SQL and SPARQL queries in order to perform effectively, if they are to remain open to having regular data updates." McRae-Spencer and Shadbolt reach the conclusion that RDF technologies are better than RDBMS strategies as in terms of scalability and realistic execution, and offer superior performance over traditional RDBMS for queries based on relationship on large-scale metadata stores because they enable timely retrieval as well as data updates.
Click Here to View Full Article
to the top


Whence Data Management?
Dr. Dobb's Journal (11/06)No. 390, P. 79; Ambler, Scott W.

The adoption of agile database methods can address a shortage of viable solutions to data quality problems among most organizations, as determined by a July Dr. Dobb's Journal poll. Though data is regarded as a corporate asset by an overwhelming majority of the organizations polled, only 40.3 percent of respondents said their organization has a validation test suite in place, while 63.3 percent of those organizations let the developers run the test suite without restrictions. Sixty percent of organizations reported having a data group, but two-thirds of respondents within those organizations said developers sometimes circumvent the data group and tackle data issues by themselves, which leads to dubious database design. Just 34.2 percent of the organizations with data groups offer data-oriented training to developers. Developer education can solve about 25 percent of the problem of developers going around data groups, while 8 percent of respondents are unaware of the data group's existence and 17 percent do not know that they should be collaborating with the data group. However, 20 percent of developers cited problems working with data professionals; 36 percent called data groups very slow to respond; and 19 percent saw little value in data groups. Teaching data management basics to developers and modern development techniques to data professionals could bridge the gap in understanding between these two groups. A hopeful sign is that 33 percent of respondents said their organizations are approaching the fixing of data sources with an evolutionary strategy.
Click Here to View Full Article
to the top


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

To unsubscribe from the ACM TechNews Early Alert Service: Please send a separate email to [email protected] with the line

signoff technews

in the body of your message.

Please note that replying directly to this message does not automatically unsubscribe you from the TechNews list.

ACM may have a different email address on file for you, so if you're unable to "unsubscribe" yourself, please direct your request to: technews-request@ acm.org

We will remove your name from the TechNews list on your behalf.

For help with technical problems, including problems with leaving the list, please write to: [email protected]

to the top

News Abstracts © 2006 Information, Inc.


© 2006 ACM, Inc. All rights reserved. ACM Privacy Policy.