Fiction-Filled Computer Code Mystery Peppered With
'Ancient' Puzzles
Pittsburgh Post-Gazette (08/16/06) Templeton, David
Carnegie Mellon University's computer science department was selected this
year to create the programming contest for the ACM SIGPLAN International
Conference on Functional Programming (ICFP) to be held in Portland, Ore.,
from Sept. 18-20. Three doctoral computer science students and one
undergraduate developed a movie-like plot that requires participants to
find clues to solve a fictive ancient codex, or ancient language, which is
in turn used to solve eight puzzles. Many teams worked around the clock
for the three days they were allotted to solve the puzzles. "Like most
competitions with no real point, people kill themselves to win it," said
Robert Harper, the CMU computer science professor who sponsored the
project. "We set the bar very high." The contest's trial was held from
July 21-24, drawing wide acclaim for its sophistication and complexity.
The storyline of the CMU contest had an ancient society in the Pittsburgh
region dedicated to the study of programming and computing. Then, several
decades ago, the construction of a local mall uncovered an indecipherable
codex that participants in the contest were tasked with unlocking to open
the ancient computer by solving the eight problems. Of the 900 registered
teams, 364 were able to open the ancient computer, but only 150 made
significant progress in solving one of the eight puzzles. "The brilliant
thing the CMU group did was put together a suite of eight problems of
varying degrees of difficulty, and I would say, as a bonus, about half the
problems had very strong connections to functional programming that
connected fairly nicely to the sponsoring organization. I'm sure the ICFP
is thrilled," said Norman Ramsey, Harvard University professor of computer
science. The winners will be announced at the ICFP next month. For more
information about ICFP, visit
http://icfp06.cs.uchicago.edu/
Click Here to View Full Article
to the top
SWAN System to Help Blind and Firefighters Navigate
Environment
Georgia Institute of Technology (08/15/06)
A team of Georgia Tech researchers is developing a wearable computing
system to help firefighters, soldiers, the visually impaired, and others
navigate unfamiliar territory, particularly when visibility is limited.
The System for Wearable Audio Navigation (SWAN), which includes a small
laptop, a tracking chip, and bone-conduction headphones that relay auditory
signals to the skull without plugging the user's ears, supplies audio clues
to the wearer to help him navigate from place to place. The Center for the
Visually Impaired in Atlanta is supplying the Georgia Tech researchers with
volunteers to test the consumer feasibility of the system through focus
groups and interviews. The idea for the project was born five years ago,
when Frank Dellaert, assistant professor in Georgia Tech's College of
Computing, met Bruce Walker, assistant professor in Georgia Tech's College
of Psychology and College of Computing, and found that their work in the
respective areas of robot location-tracking and audio interfaces could
overlap to develop assistive technologies for the visually impaired. Their
work together combines GPS-based location tracking with novel interfaces
that represent data through sonification or sound. The sensors and
tracking chip worn on the user's head transmit data to the laptop, which
then determines the user's location and the direction he is looking. It
then calculates the travel route and relays directions to the user via the
bone headphones. "SWAN consists of two types of auditory
displays--navigational beacons where the SWAN user walks directly toward
the sound, and secondary sounds indicating nearby items of possible
interest such as doors, benches, and so forth," Walker said. The next step
for the researchers is to refine the computer vision system so the SWAN can
work indoors, where GPS tracking is ineffectual.
Click Here to View Full Article
to the top
Design Automation Conference Announces Call for
Papers
Business Wire (08/14/06)
Electronic design researchers and industry professionals have until Nov.
20, 2006, to submit an original paper on new research or engineering
developments involving electronic design automation for the 44th Design
Automation Conference (DAC). Automotive Electronics will be the special
theme of DAC, and conference organizers are particularly interested in
papers on design issues and challenges involving automotive electronics.
"The size and complexity of automotive systems raises a full range of
design issues from system level integration to analog and mixed-signal
verification," says Steven P. Levitan, general chair of the 44th DAC. "We
expect this topic will lead to an exciting and diverse set of technical
presentations and panel sessions that will be of interest to a wide array
of conference attendees." There are 18 topic categories in all areas of
design automation tools and methodologies, silicon solutions, and embedded
systems, and details are available on the DAC Web site. ACM's Special
Interest Group on Design Automation (ACM/SIGDA) is a sponsor of the
conference, which is scheduled for June 4-8, 2007, at the San Diego
Convention Center in San Diego. Authors will be chosen to present their
papers during 44th DAC technical sessions, tutorials, and panels.
Click Here to View Full Article
to the top
Programmers Top Annual Salary Gains
IDG News Service (08/14/06) Weil, Nancy
The annual Enterprise Systems (ES) salary survey found across-the-board
gains in base salary for all applications and systems-related IT positions,
with applications programmers enjoying the largest pay increases. The tech
workers with the highest salaries were based in the U.S. Northeast and on
the West Coast. Leading the pack, applications programmers saw an 8.7
percent pay increase over last year, while systems administrators had the
lowest gains at 2 percent, though their annual bonuses rose more than 15
percent from last year to $3,000. Last year, only four of the seven
positions surveyed saw salary increases. This year, the survey reported
increases in each of the eight positions it evaluated: applications systems
analysts, programmer/analysts, application programmers, system programmers,
network administrators, system administrators, database administrators, and
storage administrators. ES attributes the turnaround to a generally
stronger U.S. economy. One-quarter of the respondents are employed in
government or education, 14 percent in high tech or software development, 9
percent each in finance/insurance and manufacturing, 7 percent each in
services and health care, 5 percent in utilities/transportation, and 4
percent in retail. Three-quarters of the respondents work in Windows
server environments, down 4 percent from last year. Twenty-one percent
work with mainframes, down from 28 percent last year. Forty-one percent
support at least one version of commercial Unix, and Linux environments
were steady at 31 percent, up from 30 percent last year.
Click Here to View Full Article
to the top
Subliminal Search
Technology Review (08/15/06) Sandhana, Lakshmi
While the volume of images recorded by surveillance cameras on a daily
basis far outstrips humans' capacity to analyze them all, and computer
vision systems are still too primitive to be of much use, a new brain
monitoring technology could take advantage of the brain's ability to
subconsciously detect anomalies. The brain notices anomalies, such as a
shadow where there should be none, much faster than a person can visually
or verbally identify it. Columbia University bioengineer Paul Sajda is
hoping to enable his cortically coupled computer vision system, or
C3Vision, to tap into that ability, enabling surveillance analysts to comb
through many times more images per hour than currently possible. Working
under a DARPA grant, Sajda built a prototype that consists of a string of
electrodes worn around the subject's head to monitor fluctuations in the
brain's electrical activity. The subject watches a video at 10 times the
normal speed while a computer scans the brain fluctuations for neural
signatures of unusual images or events. The system then culls out any
flagged images for further analysis. "We are aiming to speed up [visual]
search by 300 percent," Sadha said. "The system is designed not only for
finding very specific targets but also things image analysts think are
'unusual,' which is very difficult to do with a computer vision system."
Law enforcement personnel could use the system to better detect terrorist
activity, and radiologists could use the system to scan hundreds of
mammograms very quickly, picking out those that require closer scrutiny.
Photo researchers could use C3Vision to scan through images on the Web at
speeds of up to 20 per second. In testing, subjects spotted the anomaly in
a string of images flashing at 10 per second.
Click Here to View Full Article
to the top
Lawsuit Targets Pa.'s Electronic Voting Machines
Philadelphia Inquirer (08/16/06) Fitzgerald, Thomas
A group of voting-rights advocates yesterday filed suit to block
Pennsylvania from using e-voting machines that do not create a paper record
of each vote cast. The suit claims that machines that do not produce a
paper record, which are set to be used in 58 counties, are in violation of
the election code because there is no mechanism by which to audit the
results. "Whatever the initial promise may have been for electronic
voting, we now know...that they are simply not ready for prime time," said
Lowell Finley, a lawyer for the nonprofit group Voter Action. Many of the
machines that have been certified in Pennsylvania and other states have
malfunctioned, according to the suit, which cites studies that show how
easily security experts can alter vote totals, the plaintiffs claim. They
claim that the state's testing procedures are insufficient, and want
Secretary of State Pedro Cortes to decertify seven models of paperless
machines. Department of State spokeswoman Leslie Amoros counters that
there are numerous safeguards in place, such as training poll workers and
sequestering the machines. Though there must be a physical record of
ballots cast under state law, the department claims that the machines
satisfy that requirement with an auditing function that recalls the voting
histories. Amoros defended the legality of the machines, but said the
department would be willing to implement paper trails, though that
provision raises concerns that a voter's right to cast a secret ballot
could be compromised.
Click Here to View Full Article
to the top
At Media Lab, Less Whiz, More Bang
Boston Globe (08/12/06) Weisman, Robert
MIT's Media Lab, a major axis point of the digital revolution in the
1990s, is transforming itself from a blue-sky research facility focused on
multimedia and convergence to one focused on more utilitarian areas such as
health care and aging. The lab has also been entering into closer
partnerships with corporate sponsors under its new director, Frank Moss,
who emphasizes the need for new, more pragmatic approaches. Moss also
seeks to further the lab's efforts to widen access to technology for the
disabled or impoverished, such as digitally controlled prosthetic limbs and
$100 laptops. "If we direct our research at these kinds of problems, we're
setting the stage for breakthroughs that apply to everybody," Moss said.
He has launched a "buddy system" to ensure that faculty members are forging
stronger connections with the business community and developing projects
that will solve real problems. In an era when corporate research
sponsorship has plateaued, Moss has been in discussions with the Bill &
Melinda Gates Foundation and other philanthropic endeavors that might be
interested in the lab's work in areas such as health care and aging. He
also wants to consolidate the hundreds of independent research projects
into fewer groups with a broader scope. The lab is attempting to emerge
from the period following the tech bust of 2000 that saw its funding dry up
and its overseas facilities in Ireland and India close in the wake of
disputes with those countries' governments. Moss has yet to pull the plug
on any of the lab's projects, though he has held brainstorming sessions
with faculty members and students to help clarify the lab's overall
direction. Currently, some researchers are hashing out the details of
their health care projects with corporate sponsors, while others are still
pursuing far-off work on solar cars and the future of media.
Click Here to View Full Article
to the top
Forget Chess. The Real Challenge Is in the Cards
Canadian Press (08/14/06) Owram, Kristine
In 1997, IBM's Deep Blue defeated world-chess champion Gerry Kasparov, but
when it comes to poker, even the world's most powerful computers are still
bested by top human players. Skillful poker playing requires decisions to
be made based on incomplete or even inaccurate information, which computer
scientists have yet to figure out how to program. "The skills that make
human poker players really good are skills that don't seem to match well
with what computers can do," said Jonathan Schaeffer, chair of the computer
science department at the University of Alberta. "Computers aren't
particularly good at learning, for example, or reasoning by analogy."
Schaeffer helped design the Hyperborean poker-playing computer that went
undefeated at two recent tournaments hosted by the American Association of
Artificial Intelligence. In the first competition, four computers played
40,000 hands of Texas Hold 'em with a seven-second allowance to make their
move. In the second, the computers had 60 seconds to make their decision,
but only played 12,000 hands. The University of Alberta program bested its
opponents by inferring information about their hands from their decisions,
said Michael Bowling, the leader of the research group that created the
computer. The program, though perhaps the best of its breed, is still a
long way away from defeating the world's top players, but it has at least
progressed to the point where it can give them respectable competition.
Schaeffer admits that they are still a long way from developing a poker
computer of the caliber of Deep Blue. "The nice thing about chess as a
property of the game is what we call perfect information. You look at the
board, you know where all the pieces are, you know whose turn it is--you
have complete knowledge of the game," he said. "But in the real world,
knowing everything is just so rare...poker's much more representative of
what the real world's like, and in that sense it becomes a much harder
problem."
Click Here to View Full Article
to the top
Games Guru Won't Walk the Line
the age (08/15/06) Karena, Cynthia
John Buchanan is adamant about taking a different approach to exploring
media applications. "Let's do something interesting in the space before
we're consumed by linear media," said Buchanan, a professor at Carnegie
Mellon University's Entertainment Technology Center in Adelaide. "Ten
years ago linear media ruled with radio, TV, and movies. There was no
interaction with the media at all. Mobile phones, the Internet, and video
games have now replaced linear as the king of media." Buchanan warns of
the Internet becoming simply "a new way of delivering crap," and he hopes
to ensure that video games are experiential, instead of simply running
along a linear narrative that describes without engaging. Technology
should be used to create communities, Buchanan argues, pointing to
teenagers in Sweden who are creating social-networking profiles using their
Bluetooth mobile phones. He is also working toward greater adoption of
technology throughout the artistic community, which could require a more
accessible interface. Buchanan, who has served as research head for
Electronic Arts in the United States and Australia, says the success of the
video game industry has stifled innovation, because there is so much now at
stake when a new game is under development. "Developing games is now a
high-risk endeavor. The cost of prototyping becomes expensive because of
the technology needed to build it. When interacting with characters in a
video game, their behavior is scripted and hard coded. Programming the
behavior (by) anticipating for all possible scenarios makes it very
expensive." He is exploring tools that could provide players with a
reasonably priced way to test out their own ideas for video games, such as
game sketching, which brings in an actor to react to the program in real
time to refine the timing and realism of the game.
Click Here to View Full Article
to the top
E-Mail Security Hero Takes on VoIP
CNet (08/15/06) McCullagh, Declan
Phil Zimmerman, who developed free e-mail encryption software called
Pretty Good Privacy more than a decade ago, has now developed software to
secure VoIP phone calls. The software, called Zfone, has already met with
some success. A beta version released in March works with VoIP software
such as Gizmo and Free World Dialup that supports the SIP standard. But
Zimmerman's efforts to popularize Zfone have placed him at the center of a
growing political and technical debate about how to secure VoIP calls while
at the same time making it possible for police and intelligence agencies to
conduct electronic surveillance. The Bush administration, fearing that
terrorists and drug criminals will use VoIP, has demanded that broadband
Internet providers provide backdoors for government wiretapping, a
requirement that a federal appeals court ruled was permissible under the
1994 Communications Assistance for Law Enforcement Act (CALEA). Zimmerman
claims his software has no backdoors--it makes use of encryption that
scrambles the conversation from end-to-end. The lack of a backdoor could
potentially draw the ire of the U.S. government. The FBI has drafted
legislation that would require makers of networking gear to build in
backdoors for eavesdropping. If the legislation is approved by Congress,
it would prevent companies from using Zfone unless they include mandatory
surveillance backdoors for police and spy agencies.
Click Here to View Full Article
to the top
Software Speeds Up the Discovery Process Worldwide
University of Queensland (08/15/06)
Bio-scientists around the world will be able to take advantage of new
software that processes data in much less time. In addition to reducing
the hundreds of hours biomedical researchers spend selecting
high-resolution images to a fraction of that time, the rapid semi-automated
single particle selection software (SwarmPS) can be handled by a
less-skilled operator. Developed by researchers at the University of
Queensland in Australia, SwarmPS makes use of cross-correlation and
edge-detection algorithms, but what really sets the tool apart from other
technologies is its incorporation of human interaction with images to
enhance its processing power. "Essentially, SwarmPS has been designed to
provide a user-friendly, powerful and flexible graphical interface to
manage and run particle selection jobs," explains Geoffery Ericksson, a
computational scientist at the Queensland Brain Institute. SwarmPS, which
will work with most standard computer platforms, is designed to save
researchers time and money selecting key images from thousands of other
images.
Click Here to View Full Article
to the top
Weeklong Camp to EXITE Girls About IT
Computerworld Australia (08/14/06) Tay, Liz
The first of three IT camps in Australia for teenage girls offered by IBM
began in Queensland on Wednesday. The week-long Exploring Interests in IT
and Engineering (EXITE) Gold Coast Camp will attempt to show 27 secondary
school girls that careers in technology are not just for boys. "Skills
shortages in IT, especially the shortage of women entering IT, is a
significant issue for Australia if it is to remain competitive in the
global market," says Megan Dalla-Camina, strategy and marketing director
for IBM Australia-New Zealand. "EXITE Camps tackle this problem at the
grassroots level, aiming to challenge traditional perceptions about roles
in the industry." Camp participants will build and program robots, create
and design Web pages, visit Griffith University and Gold Coast Water, and
gain a female mentor from IBM who will continue to provide academic
assistance and career counseling as they return to school. A year ago the
camp drew 30 teenage girls, and 24 are now enrolled in ICT subjects and
three have worked at IBM. The company has 50 EXITE Camps worldwide
scheduled for this year, including the Ballarat Camp in mid September and
the Sydney Camp in early October.
Click Here to View Full Article
to the top
The Semantic Logger: Supporting Service Building From
Personal Context
University of Southampton (ECS) (08/11/06) Tuffield, Mischa M.; Loizou,
Antonis; Dupplaw, David
The Semantic Logger presented by a team of University of Southampton
researchers is software that supports the importing, storage, and
exploitation of personal information (metadata) through the use of Semantic
Web-enabling technologies and in compliance with as many World Wide Web
Consortium (W3C) recommendations as possible. The Semantic Logger was
designed to offer users both a public and a private knowledge base for the
purpose of determining whether any information logged was to be posted for
public consumption. The researchers report that the logger's development
was largely fueled by people's willingness to post personal information
online. The Semantic Logger system allows new services to join on an
ad-hoc basis through the virtue of its service-based architecture.
Important system components include the AKT Project's SPARQL-compliant
resource description framework (RDF) 3store, the mSpace interface, and the
Friend of a Friend (FOAF) model. The information stored in the Semantic
Logger is integrated in a pair of services outlined by the researchers: A
recommender system that creates recommendations by using any applicable
context stored in the logger, and Photocopain, a personal photo annotation
tool that blends content and context-based data. Among the information
sources tapped by Photocopain are global positioning data, camera metadata,
calendar data, image analysis, Flickr, and the Network Gazetteer. The
success of the Semantic Logger hinges on the provision of data by users.
Click Here to View Full Article
to the top
Search Me?
Washington Post (08/13/06) P. D1; Thompson, Bob
Even back when Sergey Brin and Larry Page were working out of
friend-of-a-friend Susan Wojcicki's Silicon Valley garage, they dreamt of
one day creating the world's largest digital library. Indeed, they were
supposed to be working on a digital library project when they developed the
Google search engine, according to Wojcicki, who is now a vice president at
Google. Now, eight years and hundreds of billions of dollars later, Brin
and Page are trying to make their dream a reality. Under the Google Book
Search program, Google has partnered with six libraries to digitize their
collections, but five major publishers and the Authors Guild have filed
copyright infringement suits seeking an injunction to halt Google's
scanning. Google claims that its scanning program falls within the
boundaries of "fair use," and that it will only offer searchers a snippet
from the work. Stanford University head librarian Bill Keller is
enthusiastic about the project because he fears an uncertain future for
printed materials. Google's partner libraries deal with the copyright
issue in different ways. Oxford and the New York Public Library are only
allowing Google to copy works in the public domain. The University of
Michigan is taking the opposite approach and opening its entire collection
to Google, as is the University of California, which just signed on to the
program last week. Harvard agreed to a limited test program at first, and
Stanford is only allowing Google to scan out-of-print titles, at least
until the lawsuits are settled. Initially, Google had negotiated with
publishers for the right to scan their in-print books, but then Google
announced its library partnerships and they began to get nervous. Google
has long had an "opt-out" policy for content owners who do not want their
books included in the database, but publishers say that is not enough, and
that they should be in control and that the program should be governed by
an opt-in policy. Google likens the scanning project to basic Web search,
where site operators can opt out of having their page displayed in results
listings. The publishers counter that books, unlike most Web sites, are
not designed to be available for free.
Click Here to View Full Article
to the top
DNA Processors Cash in on Silicon's Weaknesses
New Scientist (08/04/06) Vol. 191, No. 2563, P. 24; Simonite, Tom
The notion that the same tools that scientists use to manipulate DNA
strands could provide the computational muscle for a computer, despite
having emerged with such promise, faded quickly as scientists realized that
they could not control the behavior of DNA the same way they can electrons
in conventional computers. "Only 90 percent of our molecules would do what
they were supposed to," said University of Southern California computer
scientist Leonard Adleman, who first solved a mathematics problem with DNA
molecules in a test tube 12 years ago. "That was the limiting factor when
it came to building bigger, faster systems." Since 2002, many scientists
have been looking to DNA more for building nanostructures than for
computers. Interest has not faded entirely, however, as a small group of
dedicated researchers is still exploring DNA computing--not so much for
crunching numbers, but for probing biological systems. Biological hardware
could significantly accelerate the identification of viruses and genetic
markers that point to diseases-related genes. A growing number of
biologists are using microarray chips with DNA fragments to identify
pathogens, though the technology is hampered by its slowness, while a DNA
computer could perform the same task much faster, says Columbia University
virologist Joanne Macdonald. Macdonald is using the logic gates of a DNA
computer called MAYA, built by Columbia's Milan Stojanovic, to produce a
computational tool that she claims can detect specific sequences within 15
minutes. Using the logic gates, Macdonald has already determined the
difference between different strains of West Nile virus. Other researchers
are still looking to DNA computing to perform basic biological
calculations, such as processing human genes to look for disease
indicators.
Click Here to View Full Article
- Web Link to Publication Homepage
to the top
What's the Greatest Software Ever Written?
InformationWeek (08/14/06)No. 1101, P. 39; Babcock, Charles
Charles Babcock evaluates what he considers to be the greatest software
programs ever written based on such wide-ranging criteria as historical
context, real-world adoption, and social impact. He rates
Unix--specifically, Berkeley Software Distribution (BSD) 4.3--as the single
greatest piece of software ever created in view of its effects on the
world. The software traces its roots to Bell Labs researcher Ken
Thompson's Uniplexed Information and Computing System (Unics), which
facilitated the simultaneous use of a computer by two people, and was
rechristened Unix with the addition of text formatting. Bill Joy and other
researchers would add to Unix, and their extensions were compiled into BSD,
of which version 4.3 is "the single biggest undergirder of the Internet,"
according to Babcock. The second greatest piece of software is IBM's
System R, the root architecture of the relational database, while the third
greatest piece of software is the gene-sequencing software at the Institute
for Genomic Research, which is credited for "accelerating the science of
genomics by at least a decade," according to venture capitalist Gary
Morgenthaler. Other breakthroughs cited by Babcock include the MIT
Instrumentation Lab's Apollo spacecraft guidance system, which could
operate on a tiny amount of available memory; Google's page-ranking search
application; the Java language, which made the use of intermediate byte
code fashionable; the IBM System 360 operating system, which introduced the
concurrent operation of different applications on one computer system; the
Morris worm, which demonstrated the hazards of increasing
interconnectedness and the vulnerability of the Internet; the Mosaic
browser, which brought what Babcock terms "a fresh technical synthesis" by
combining address lines, mouse-based pointing and clicking, multimedia file
displays, and hyperlinking in windows; and American Airlines' Sabre system,
which showed that software could bridge the gap between tactical and
strategic business needs.
Click Here to View Full Article
to the top
The Invisible Assistant
Queue (08/06) Vol. 4, No. 6, P. 44; Borriello, Gaetano
University of Washington researchers started to investigate issues facing
the development of computers that can assist people by anticipating their
needs with Labscape, a ubiquitous computing application/environment, writes
University of Washington computer science professor Gaetano Borriello. The
creation of Labscape began with the organization of a framework for
assembling data associated with an experiment by specifying a flow graph
that could classify all the steps of an lab experiment into eight basic
categories. A researcher's location and activities during the experiment
were logged through various sensors, generating context awareness. Upon
completion of an experiment, the data used to document its steps and
results was automatically compared to the original experiment flow graph.
The application offered researchers' colleagues, who may have helped with
some steps, the convenience of not having to concern themselves with
transferring information. Keeping the data and experiment design in a
machine-readable format allowed the greater research community to run
searches on experiments based on similar samples, processes, or results,
thus making it easier to duplicate the results of earlier experiments or to
formulate the next steps to follow in their own experiments. Among the
lessons the Labscape team took away from the experience was the need for a
visible user interface that can check the system's operational performance
as well as override its proactive functions; the logic of considering
sensor fusion upfront in order to address ambiguity; the advantages of
incremental system installation; the need to standardize data formats to
develop a supporting tool infrastructure; the wisdom of keeping work
practices consistent at all times, and only effecting changes within
reason; and the need for fail-safes.
Click Here to View Full Article
to the top