Google Calls for International Standards on Internet
Privacy
Washington Post (09/15/07) P. D1; Rampell, Catherine
Speaking before a U.N. audience in Strasbourg, France, global privacy
counsel for Google Peter Fleischer said that fragmentary international
privacy laws burden companies and fail to protect consumers, arguing for
new international standards on the collection and use of consumer data.
Fleischer said the United Nations should create standards countries could
adopt and adjust to fit their needs. "The ultimate goal should be to
create minimum standards of privacy protection that meet the expectations
and demands of consumers, businesses, and governments," Fleischer said.
Google has frequently been criticized for its privacy policies and is
currently under investigation by the European Union for violating global
privacy standards. Critics are also concerned that Google's planned $3.1
billion merger with online advertising broker DoubleClick would place too
much consumer data in the hands of one company. Fleischer criticized U.S.
privacy law for being too complex and too patched-together because
different laws apply to different industries and vary by state. Fleischer
also called the European Union model "too bureaucratic and inflexible."
Fleischer suggested adopting a model similar to the Asia-Pacific Economic
Cooperation guidelines, but critics say the APEC standards are too lenient.
"The APEC guidelines are far below what Google would be expected to do in
Europe or the United States," said Electronic Privacy Information Center
executive director Mar Rotenberg. APEC does not limit data collects, for
example, which is a significant problem and the key point in the dispute
over Google's business practices, Rotenberg said.
Click Here to View Full Article
to the top
Robots That Sense Before They Touch
Technology Review (09/17/07) Greene, Kate
Intel researchers have developed a sensor for robotic arms that allows the
robot to sense an object before actually touching it. For example, a
robotic arm is capable of telling the difference between an empty bottle
and a bottle filled with water without touching either one. The
technology, known as pre-touch, is intended to "improve the ability of
robots to grasp objects in unstructured human environments," says Intel
research scientist Josh Smith. Pre-touch's electric-field (EF) proximity
sensors are electrodes made of copper and aluminum foil. A current is sent
to one of the electrodes, which creates a magnetic field and induces a
current in the other electrodes. When the robotic hand gets near metal or
anything with water in it, the object reduces the induced current, which is
detected by the sensors. Special algorithms process the information and
instruct the robotic hand to move around the object accordingly. Smith
developed similar EF sensors while he was a student at MIT. Those sensors
were used to help determine the position of people in a car, information
that was used to determine how to deploy airbags during an accident. Smith
says his current EF research will now involve developing algorithms capable
of handling the complex data that EF sensors produce, particularly when the
object or the robot is in motion.
Click Here to View Full Article
to the top
CHI 2008 Conference
Designophy (09/14/07)
The CHI 2008 Conference, sponsored by SIGCHI, ACM's Special Interest Group
on Computer-Human Interaction, offers the CHI community an opportunity to
share information and ideas through its Papers venue. Through Sept. 19,
2007, the CHI community can submit case studies on the development and
application of interactive systems that will significantly impact our
understanding of human-computer interaction. Interaction technologies
should focus on a new technique or device, while reflective analyses should
offer full support of conclusions made about HCI issues. CHI is also
welcoming other papers on interactive systems, including those that
concentrate on methods and tools, theories and models, as well as fieldwork
and ethnography reports, and laboratory reports. Authors will have an
opportunity to present their paper in a 25-minute talk, including
questions, and CHI will notify participants on Dec. 10, 2007. The
conference is scheduled for April 5-10, 2008, in Florence, Italy. CHI 2008
will offer a wide range of venues from workshops to doctoral consortium to
courses to technical programs, in addition to social activities.
Click Here to View Full Article
to the top
Weight Loss Game Looking for 'Neat-O' Results
University of Houston News (09/12/07) Holdsworth, Ann
University of Houston computer science professor Ioannis Pavlidis and
research assistants Yuichi Fujiki and Kostas Kazakos have developed a
computer game targeted at overweight people that uses motion sensitivity to
encourage mild exercise. Non-Exercise Activity Thermogenesis (NEAT-o)
games, including races and logic puzzles, can be played on any handheld
personal digital assistant while the user wears a sensor that detects
movement such as running, walking, bending over, and even foot tapping.
The movement data is wirelessly sent to the PDA, where the player can see
the game avatar move in real time. The game can run on the PDA all day to
track the user's movements as he or she walks around the office or home,
and can be connected to other handheld devices to compete against other PDA
users. Mayo Clinic physician and leading authority on obesity James Levine
says a lack of daily mild exercise is largely responsible for the world's
obesity epidemic. Levine created the term "NEAT" to cover any physical
activity that is not conscious exercise. Pavlidis hopes the NEAT-o games
will increase physical activity, add a little bit of fun, and build the
NEAT mindset into modern lifestyles. "The allure of computer gaming and
competition with other users encourages players to make small lifestyle
changes that can add up to big health benefits," Pavlidis says.
Click Here to View Full Article
to the top
Technique Links Words to Signing
BBC News (09/15/07) Adams-Spink, Geoff
IBM researchers have developed SiSi, a system that translates spoken words
into British Sign Language (BSL). SiSi, short for say it sign it, enables
deaf people to have simultaneous sign language interpretations whenever a
human interpreter is not available, and could possibly be used for signing
for television, radio, and telephone calls. SiSi uses speech recognition
to animate a digital character. SiSi has already been approved by the
Royal National Institute for Deaf people (RNID). "RNID welcomes any
development that would make the information society a more equal place for
deaf and hard of hearing people," says RNID's Guido Gybels. SiSi was
developed by students during a 12-week initiative called Extreme Blue that
IBM hosts. "We had a profoundly deaf mentor, so he kept a close eye on
what was being done and checking whether our translation corresponded to
real BSL," says Maria Vihljajeva, a student who developed the business plan
for SiSi. In addition to the BSL avatar, the students also developed an
avatar that uses Sign Supported English, a more direct translation that
uses conventional syntax and grammar. Both avatars were developed by the
University of East Anglia. Tom Klapiscak, another student who worked on
the project, says SiSi was designed so other languages could be added
fairly easily, though creating the translation module would be a fair
amount of work.
Click Here to View Full Article
to the top
New Entanglement Findings Advance Quantum Computing
Research
University of Maryland (09/07/07) Tune, Lee
University of Maryland and University of Michigan physicists demonstrated
quantum "entanglement" between a pair of totally disconnected individual
atoms a meter apart in separate enclosures through the careful manipulation
of photons the atoms emitted, and the work could be an important step
toward the creation of incredibly fast quantum computers. "Now that this
technique has been demonstrated, it should be possible to scale it up to
networks of many interconnected components that will eventually be
necessary for quantum information processing," says physicist and research
team leader Christopher Monroe. A second area of investigation by
University of Maryland researchers focuses on the use of solid-state
electronic devices to reach quantum states. Four years ago physicists
probed the existence of entangled states between two quantum bits (qubits)
produced with a solid state circuit called a Josephson junction. Qubits
can exist in multiple states simultaneously through the phenomenon of
superposition, and thus can exponentially accelerate problem-solving when
arranged into a computer. The fragility of entangled quantum
superpositions is a major challenge, and Monroe says his team's experiment
"shows for the first time that it is possible to use emitted photons to
entangle the fixed atoms that emitted them." He notes that while atoms are
ideal for serving as qubits, photons are ideal tools for the transference
and control of data. The National Security Agency, the Disruptive
Technology Office, and the National Science Foundation Information
Technology Research and Physics at the Information Frontier programs
supported the work of Monroe's team.
Click Here to View Full Article
to the top
NSF Researchers Produce RFID Random Number
Generator
Government Computer News (09/12/07) Jackson, Joab
University of Massachusetts researchers, with funding from the National
Science Foundation, have developed an inexpensive way of producing truly
random numbers for radio frequency identification tags, as well as a
technique that produces a unique fingerprint for each tag. Although
encryption programs require a reliable source of random numbers, computers
are incapable of producing truly random numbers. Algorithms have been
developed that can help machines produce numbers that statistically
resemble random numbers, but they contain subtle repeatable patterns that
can be used to decipher a message encrypted with those digits. The
technique developed by the researchers produces a set of random numbers
from an RFID tag by reading the binary states of the tag's memory cells
while the tag is being powered up. A typical Electronic Product Code Class
1 RFID tag has between 1,000 to 4,000 gates, which is typically volatile
memory that loses all information when power is lost. Each time a tag is
powered up, a certain number of gates fluctuate randomly between having a
residual charge or not having a charge. These fluctuations can be used to
produce a stream of random numbers. The researchers say the numbers
produced by this process have passed the National Institute of Standards
and Technology's test for statistical randomness. The variations in each
tag's gates can also be used to uniquely identify each tag, ensuring
information derived from each tag has not been altered by a possibly
malicious source.
Click Here to View Full Article
to the top
Software Transforms Digital Photos Into Old-Fashioned
Paintings
PhysOrg.com (09/12/07) Zyga, Lisa
Dalhousie University computer scientist Stephen Brooks has developed
software capable of turning digital images into mixed-media paintings and
drawings with almost no user input. Brook's system analyzes the image and
creates a measure of detail for each pixel. Areas with different levels of
detail are divided into regions, with each region processed independently
with a different "nonphotorealistic rendering," which can include paint
dubs, soft glow, crystallize, ink outlines, and cartoon. Selecting the
filters is the only step of the process that requires instruction from the
user. The different areas are blended together, which can either be done
manually by the user or automatically by the system. To be able to convert
facial portraits, Brooks had to develop systems for skin filtering,
face-point detection, and face-point clustering. "The most advanced
component of my system is the module that handles portraits," Brooks says.
"Automatic face detection and processing is quite tricky if one wants to
make the system robust. I use a combination of skin detection, cascaded
weak detectors, and Gaussian mixture model clustering." Brooks is
currently developing more advanced features, including integrating
2.5-dimensional physical materials such as felts and cardboards, and
possibly adapting the technique for videos.
Click Here to View Full Article
to the top
Microformats Hop on Semantic Web 'Griddle'
InternetNews.com (09/12/07) Kerner, Sean Michael
Microformats and Gleaning Resource Descriptions from Dialects of Languages
(GRDDL) Semantic Web construct both use metadata descriptors to define
content, but use slightly different approaches. The World Wide Web
Consortium (W3C) wants to bridge the gap between the two approaches.
"Microformats and the Semantic Web always had a lot in common at a high
level, but GRDDL fills in a few technical details so that they interoperate
at a practical, running-code level," says Dan Connolly, the W3C GRDDL
Working Group staff contact. Microformats use XHTML and HTML while GRDDL
uses Resource Description Framework data from XML so it can be transformed
and understood by other application in a mashup or application setting.
"Microformat authors that want their data to integrate seamlessly with
other semantic Web data should use well-formed XHTML and profile URIs,"
Connolly says. "For example, the dbpedia project takes millions of facts
from Wikipedia and exports them using URIs and RDF and SPARQL." Connolly
says notes that HTML also has a special profile hook to use a URI to share
document-local conventions on the Web. Some microformat documents use the
hook, but GRDDL links the profile documents to XSLT transformations so
microformats and other syntax conventions can be exploited.
Click Here to View Full Article
to the top
Researchers Improve Ability to Write and Store
Information on Electronic Devices
Argonne National Laboratories (09/13/07) Carson, Sylvia
Argonne National Laboratory physicist Matthias Bode has demonstrated a
technique that makes it possible to switch a magnetic nanoparticle without
the use of a magnetic field, which could lead to more accurate and
efficient data storage. A special scanning tunneling microscope (STM) was
used to force a spin current through a small magnetic structure. Bode and
his colleagues demonstrated that the structure's magnetization direction is
not changed by a small current, but can be manipulated if the spin current
is high enough. Bode's experiment focused on magneto-resistive random
access memory (MRAM), which uses magnetic storage elements consisting of
two ferromagnetic layers separated by a thin non-magnetic layer. MRAM
stores data by applying an external magnetic field to one of the layers to
polarize it while the other layer remains polarized in a constant
direction. As memory devices continue to become smaller, they are more
susceptible to "false writes" or "far-field" effects, Bode says, where the
magnetic field switches the magnetization of too many bits. Using a STM,
Bode was able to manipulate a bit without altering any others accidentally,
eliminating the false write effect. STMs could also allow scientists to
look for small impurities in magnetic storage structures and to see how
impurities affect the magnet's polarization, possibly leading to new
materials or methods that make memory storage more efficient.
Click Here to View Full Article
to the top
Leading-Edge Body Sensor Could Help Produce Sporting
Champions
Engineering & Physical Sciences Research Council (09/13/07)
Imperial College London researchers are developing a sensor that can be
worn behind the ear to collect information on posture, stride length, step
frequency, acceleration, and response to shock waves during athletic
training sessions. The information can be automatically transmitted to a
handheld device or laptop used by a coach, who can use the data to create
on-the-spot advice and instructions, allowing for ultra-effective training
sessions. "The sensor we're working on is inspired by the semi-circular
canals of the inner ear, which play a key role in controlling our motion
and balance," says professor and project leader Guang Zhong Yang, a leading
body sensor networks researcher. The data generated provides an authentic
and realistic indication of how the athlete would perform without the
sensor, which is often not the case for other body sensors because they can
be cumbersome and cause poor performance, Yang says. The sensor could also
be used to monitor patients suffering from a variety of injuries or
illnesses, and could help promote a healthy lifestyle. The sensor could
monitor patients with degenerative arthritis or neurological gait
abnormalities. The device could also be used to create movement-based
computer games and virtual reality-based sports training.
Click Here to View Full Article
to the top
University Program Targets Online Security
Augusta Chronicle (GA) (09/12/07) Few, Jenel
Armstrong Atlantic State University is researching and developing software
that will be able to intercept secret messages transmitted over the
Internet and destroy any malicious content. As part of a demonstration of
the program, professor Ray Hashemi showed how a message can be hidden in a
photograph and go unnoticed by the human eye. "A terrorist headquarters
can send a hidden message to a sleeper cell in this photograph or music or
the text of an email," Hashemi says. "What we did was develop a vaccine
that is able to intercept the image; the sleeper cell will get the image
with the message destroyed and the original can be held for a government
agency to analyze later." The university's School of Computing is
developing the program for companies that provided its new Cyber Security
Research Institute with equipment and other resources. The institute also
demonstrated other top-secret projects that are unlikely to receive further
public attention due to the sensitive nature of the work.
Click Here to View Full Article
to the top
Computerized Treatment of Manuscripts
Universitat Autonoma de Barcelona (09/06/07)
Researchers at the UAB Computer Vision Center have developed a more
reliable computerized system for recognizing the content of difficult to
read manuscripts, handwritten scores, and architectural drawings. The BSM
(Blurred Shape Model) system is based on the way in which the human mind
sees and interprets images to describe and classify models of handwritten
symbols. The tool is unique in that it is able to detect variations,
elastic formations, and uneven distortions when letters, signs, drawings,
and any other type of symbol is manually reproduced, and it has the
potential to work in real time, which would be a few seconds after a
document has been entered into the computer. BSM would serve as a human
machine interface for automatically reproducing documents as they are
written or drawn. Current systems for reading handwritten symbols use the
same process for detecting different types of symbols, while BSM is able to
adapt to different areas, using a grid for dividing images into sub-regions
to analyze and recognize symbols. In tests, BSM bested similar systems
with more than 98 percent exactness in recognizing 18th and 19th century
musical scores, and 90 percent exactness for recognizing handwritten
architectural symbols.
Click Here to View Full Article
to the top
Internet Governance Forum to Meet in Brazil, 12-15
November
eGov Monitor (09/10/07)
The second Internet Governance Forum will be held in Rio de Janeiro from
Nov. 12 to Nov. 15. Conference organizers expect about 2,000 attendees
from over 100 countries to attend, including representatives from
government and non-governmental organizations, members of the private
sector, and the Internet community. Last year's meeting in Athens centered
on issues of access, diversity, openness, and security on the Internet. In
addition to these themes, this year's meeting will touch on the use and
direction of critical Internet resources, including administration of the
Domain Name System and Internet Protocol addresses, technical standards,
and administration of the root server system. Although not a
decision-making body, the Internet Governance Forum will seek to promote
dialog on public policy issues related to the Internet and provide
opportunities for participating organizations to work together.
Click Here to View Full Article
to the top
Quantum Threat to Our Secret Data
New Scientist (09/13/07) Vol. 195, No. 2621, P. 30; Graham-Rowe, Duncan
Quantum computing's ability to decrypt the codes that safeguard banking,
e-commerce, and business data has taken a step closer to realization with
the development of quantum computers that can run Shor's algorithm by two
research groups working independently. RSA is an example of a highly
common encryption system that can be defeated by Shor's algorithm. RSA
involves a widely distributed public key for encrypting messages and a
secret private key for decrypting them, and the trick to solving the
private key is to work out the large prime numbers that produce the key
when they are multiplied together. Shor's algorithm dramatically reduces
the time it takes to find the prime factors by searching for telltale
patterns in remainders when a key is divided by a prime factor, but quantum
computation is essential for performing the massive number of calculations
that the algorithm requires to be successful. The first quantum
implementation of Shor's algorithm involved the manipulation of nuclear
spin, while the more recent experiments--one led by Andrew White at
Australia's University of Queensland and the other by Chao-Yang Lu of the
University of Science and Technology of China--used quantum photonic
computers. Photon pairs were produced with femtosecond lasers and passed
through polarizing bismuth borate crystals to generate entangled qubits,
which were coaxed by optical devices to run Shor's algorithm to factor the
number 15 into its prime components. IT security specialist Bruce Schneier
calls the development of techniques to run the algorithm using standard lab
optics a significant achievement that could spell trouble for encryption
down the road. White contends that cryptography would need to be
fundamentally rethought if quantum calculations for much larger numbers
could be carried out. "There are paths to a fully scalable quantum
computer," he notes.
Click Here to View Full Article
- Web Link May Require Paid Subscription
to the top
Checkers Is Solved
Science (09/14/07) Vol. 317, No. 5844, P. 1518; Schaeffer, Jonathan;
Burch, Neil; Bjornsson, Yngvi
As many as 200 computers running since 1989 have applied artificial
intelligence methods to consider the approximately 500 billion billion
possible positions for the game of checkers, and determined that a draw is
the inevitable outcome of perfect play. "Perhaps the biggest contribution
of applying AI technology to developing game playing programs was the
realization that a search-intensive ('brute-force') approach could produce
high-quality performance using minimal application-dependent knowledge,"
the authors write. "Over the past two decades, powerful search techniques
have been developed and successfully applied to problems such as
optimization, planning, and bioinformatics. The checkers proof extends
this approach by developing a program that has little need for
application-dependent knowledge and is almost completely reliant on
search." The solving of checkers has a high dependence on endgame
databases, which are compiled via computations from the end of the game
back toward the starting position. Forward search is conducted through the
maintenance and exploration of a tree of the proof in progress to generate
positions that must be probed in order to further the proof's progress.
The third algorithm/data component of the proof procedure is the proof
solver, which employs two programs to ascertain the value of a given
position. The first program produces a heuristic value for the position,
while the second program utilizes a space-efficient modification of Proof
Number search to yield a proven, partially proven, or unknown result. The
authors conclude that the checkers proof raises the bar in terms of what is
achievable by search-intensive algorithms, and "provides compelling
evidence of the power of limited-knowledge approaches to artificial
intelligence." They note that the discovery of knowledge is an implicit
function of deep search, and this puts search algorithms in a prime
position to exploit growth in on-chip parallelism that will soon be offered
by multicore computing.
Click Here to View Full Article
- Web Link May Require Paid Subscription
to the top
Services at UCSC
CITRIS Newsletter (08/07)
Knowledge Services and Enterprise Management (KSEM) is a one-year,
CITRIS-supported program offered by UC Santa Cruz that "[gives] students
the skills to address challenges faced in high-tech enterprises that
require an integrated understanding of both technology and business to
solve complex problems," says Information Technologies Institute director
Patrick Mantey. Economists reckon that services currently constitute up to
80 percent of the American economy, and a major portion of that percentage
consists of "knowledge services." Each instance of service delivery offers
providers an opportunity to learn something of value from the interaction,
and UCSC professor Ram Akella says learning and incorporating new knowledge
into operations is critical to businesses' sustained success. Mantey
describes KSEM as a hybrid program that cross-pollinates business
management and computer science and IT, and directs the combined discipline
toward the influence of the global knowledge-based economy. "Students
learn some elements of marketing, sales, finance, product development, and
about the supply chain," Akella explains. "They learn how to integrate all
those in a universe where we are constantly getting information." Akella
expects every successful future enterprise to have an information
management component, and two central goals of KSEM are giving businesses
the know-how to compete via solid information management and arming
proficient engineers and businesses with that knowledge. Mantey and Akella
have witnessed the growth of a national consensus around the importance of
knowledge sciences and services in lockstep with the KSEM program's
expansion, and both the National Academy of Sciences and the National
Science Foundation have emphasized the value of knowledge sciences
investment.
Click Here to View Full Article
to the top