N.J. Voting Technology in Question After Discrepancies in
February Vote
Daily Princetonian (04/14/08) Wolff, Josephine
An investigation into discrepancies recorded by several of New Jersey's
electronic voting machines shows that some of the tallies from February's
primary elections may not add up. On April 8th, a court subpoenaed
electronic voting machines used for the primary elections in six New Jersey
counties, questioning the accuracy and security of the machines. Later
that day, Sequoia Voting Systems, the manufacturer of the machines, filed a
motion to suppress the subpoenas, arguing that the subpoenas sought to test
their machines under "unknown circumstances and protocols," which could
unfairly undermine the reputation of Sequoia's machines and the public's
confidence in election results. The controversy surrounding the New Jersey
primaries started in March, when a Union County clerk noticed that the
number of Democratic and Republican voters recorded by the DRE paper
reports generated after the election did not match the number of votes cast
in each primary on those machines. For example, one machine recorded that
60 Republican and 362 Democrat ballots were activated, but 61 votes were
cast for Republican candidates and only 361 were cast for Democrats.
Princeton University professor Ed Felten says it is not the size of
discrepancy that is alarming, but that a single machine is disagreeing with
itself on how many voters voted. Similar discrepancies have been
discovered on at least eight other machines in Union County and several
more throughout the state. Sequoia has issued a memo blaming the
discrepancies on New Jersey poll workers.
Click Here to View Full Article
to the top
Robots, Our New Friends Electric?
Guardian Unlimited (UK) (04/14/08) Jha, Alok
Engineers, psychologists, and computer scientists from across Europe this
week will start Lirec, a European Union-backed project that aims to develop
the first robot personalities. "What we're looking at here is long-term
interactions between people and robots in real situations," says Lirec
coordinator Peter McOwan of Queen Mary, University of London. "The big
question is: What sort of properties does a synthetic companion need to
have so that you feel you want to engage in a relationship with it over an
extended period of time?" Lirec (Living with Robots and Interactive
Companions) is a collaborative project between 10 university partners from
seven countries that will run for just over four years. In the future,
McOwan imagines robots that help around the house and act as companions,
capable of tasks such as ordering groceries online. Robotic personalities
could also be used to assist the elderly. A concept called "spirit of the
house" would use an artificial intelligence with personality to make sure
elderly residents have not fallen or forgotten to take their pills.
University of Hertfordshire artificial intelligence professor Kerstin
Dautenhahn has developed a robot in the shape of a two-year-old boy capable
of making facial expressions and playing simple games. Dautenhahn created
a study in which a home-help robot interacts with volunteers so researchers
can study how people make longer-term relationships with machines. The
study has found that the look of the robot should depend on the person it
mostly interacts with, with extroverts preferring humanoid robots and
introverts preferring more mechanical robots.
Click Here to View Full Article
to the top
He Wrote 200,000 Books (But Computers Did Some of the
Work)
New York Times (04/14/08) P. C1; Cohen, Noam
Insead professor Philip M. Parker has developed computer algorithms that
collect publicly available information on a subject and, with the help of
60 to 70 computers and six or seven programmers, combine the results into
books in a variety of genres, printing them only when a customer buys one.
Parker has generated more than 200,000 books using his technique, making
him "the most published author in the history of the planet," he says.
Parker says medical libraries collect nearly everything he produces. He
has expanded the technique to crossword puzzles, rudimentary poetry, and
scripts for animated game shows. Parker admits that his books are
essentially worthless to someone who is good at using the Internet, but
that there are people who are not Internet savvy who find the books useful.
Artificial intelligence researchers say computers are far from being able
to substitute for what the general public would consider authors. Rutgers
computer science professor Chung-chieh Shan says being able to write a text
with the variety one would expect from a typical human English speaker is
actually the holy grail of computer linguistics, and Parker's program falls
somewhere between there and a more simple program capable of automatically
writing a telephone directory.
Click Here to View Full Article
to the top
A Field Guide to Genetic Programming
Press Release (04/14/08) Poli, Riccardo; Langdon, William B.; McPhee,
Nicholas Freitag
The automated programming technique known as genetic programming (GP) is
the subject of "A Field Guide to Genetic Programming," a new book by
Riccardo Poli, William B. Langdon, and Nicholas Freitag McPhee. GP is a
systematic, domain-independent method that computers can use to solve
problems automatically, starting from a high-level statement of the task,
says Poli, a professor in the Department of Computing and Electronic
Systems at the University of Essex. The technique makes use of ideas from
natural evolution, initially starting from random computer programs, then
refines them through processes that are similar to mutation and sexual
recombination, and the result is high-fitness solutions. Users do not have
to know or specify the form or structure of the solution in advance, Poli
says. Novel scientific discoveries and patentable inventions have already
emerged from GP. The book is available for free under a Creative Commons
license.
Click Here to View Full Article
to the top
Google Aims to Penetrate the Deep Web With HTML Forms
Crawling
Computerworld (04/11/08) Havenstein, Heather
Google has been using HTML forms such as drop-down boxes and select menus
to find Web pages in the "Deep Web," content that is otherwise invisible to
search engines. The company sees HTML forms crawling as another way to
improve its coverage of the Web, and says the ability to lead users to
documents in the Deep Web ultimately will enhance the search experience.
With text boxes, computers automatically select words from the site that
has the form; and for select menus, check boxes, and radio buttons on the
form, the crawling and indexing team selects from among the values of the
HTML. "Having chosen the values for each input, we generate and then try
to crawl URLs that correspond to a possible query a user may have made,"
say Google's Jayant Madhavan and Alon Halevy in a blog post. "If we
ascertain that the Web page resulting from our query is valid, interesting,
and includes content not in our index, we may include it in our index much
as we would include any other Web page." The team does not engage sites
that include instructions against crawling, and omits forms that require
passwords or personal information. The method does not affect page
ranking.
Click Here to View Full Article
to the top
DHS Offers First Take on Cyber Storm Exercise
IDG News Service (04/10/08) McMillan, Robert
Roughly 2,500 emergency response managers from government agencies and
companies in the United States, Britain, Canada, Australia, and New Zealand
took part in Cyber Storm II, a week-long cybersecurity simulation held by
the U.S. Department of Homeland Security. The simulation featured several
mock attacks on computer and transportation systems. In one of the
simulated attacks, a disgruntled employee of a chemical company sabotaged
his former employer's computer network. In addition to responding to the
attack, participants also had to monitor the simulated media responses to
the incident and determine the level of hearsay in news reports. Dow
Chemical's Christine Adams says the simulation allowed emergency response
managers to determine whether their plans worked as they expected them to,
and if people responded to an incident in the manner that planners expected
them to. "You think you know how people are going to respond ... but they
surprise you sometimes," says Michigan CISO Daniel Lohrmann, a Cyber Storm
II participant. The DHS is not planning to release specific information
about the simulation until it releases the after-report, which is expected
in August at the earliest. Another simulation, dubbed Cyber Storm III, is
being planned for 2010.
Click Here to View Full Article
to the top
CMU at Forefront in Building Thinking Machines
Pittsburgh Tribune-Review (04/06/08) Houser, Mark
Carnegie Mellon University is at the vanguard of efforts to create
intelligent machines, and one CMU initiative focuses on building robots
that can work in teams, with the goal of trouncing the world's champion
soccer team by 2050. Other CMU artificial intelligence projects include
driverless vehicles, computer vision, speech recognition, imbuing computers
with creativity, and brain simulation. CMU professor Alexei Efros is
expected to introduce a computer program that guesses where a photo was
shot by comparing the image to millions of others posted on an
Internet-sharing site where users mark their images with location
coordinates. A vastly enhanced Google-like search engine that not only
looks for matching words on Web pages but also reads and understands the
pages and drafts a summary is the brainchild of researchers Jaime Carbonell
and Anatole Gershman. Meanwhile, CMU's Tom Mitchell and Marcel Just earned
a private grant of $1.1 million to map neurons in the human brain that are
stimulated by thoughts about individual objects, with the aim of mapping
the neural pattern for all English common and abstract nouns. "My guess is
that there's nothing in principle that makes it impossible for computers to
be intelligent," Mitchell says. "It's just our own stupidity and our
inability thus far to figure out how to do it."
Click Here to View Full Article
to the top
Sun Labs Open House Highlights Community Contribution to
Research Projects
HPC Wire (04/10/08)
Sun Microsystems Laboratories' recent open house included demonstrations
of the lab's current research projects, including Project Wonderland, a
toolkit for building 3D virtual worlds, and Project Fortress, a new
high-performance computing programming language. "Sun Labs has delivered
measurable and results-oriented innovation through its history and by
increasingly adding collaboration to our innovation, we're taking the our
research projects even further," says Sun Labs director Bob Sproull. "Our
open-sourced prototypes are effectively building communities of research
partners, inviting participation and disruptive thinking from developers
outside of Sun Labs and the broader Sun community." Other research
projects underway at Sun include Project Darkstar, a Java-based software
infrastructure platform designed to simplify the development and operation
of massively scalable online games, virtual worlds, and social networking
applications; and Project Caroline, which includes research into
technologies for rapid and efficient development and delivery of
dynamically scalable Internet-based services. Other projects include
Project Sun SPOT (Small Programmable Object Technology), a Java-based
platform for developing wireless sensor, robotics, and swarm intelligence
applications.
Click Here to View Full Article
to the top
Tai-Chi - Future Interactive Interface
The Future of Things (04/10/08) Rattner, Ehud
European human-computer interaction researchers are trying to create
tangible interfaces that will make computer interaction possible via
augmented physical surfaces, graspable objects, and ambient media such as
walls and tabletops. The goal is to make interaction natural and eliminate
the need for a handheld device. The Tangible Acoustic Interfaces for
Computer Human Interaction (Tai-Chi) Project is currently researching
acoustics-based remote-sensing technologies with the goal of transferring
information pertaining to an interaction by using the structure of the
object as the transmission channel. Possible applications could include
wall-sized touch panels, three-dimensional interfaces, and robust
interactive screens capable of withstanding harsh environments. The
project's goal is to develop acoustics-based remote-sensing technology that
could be adapted to physical objects to create tangible interfaces.
Tai-Chi project scientists say users will be able to communicate freely
with a computer, interactive system, or the cyber-world using ordinary,
everyday objects. The project is developing different methods for
contact-point localization. Efforts include utilizing the
location-signature embedded in the acoustic wave patterns caused by
content, and triangulation and acoustic holography.
Click Here to View Full Article
to the top
Software Program Helps the Disabled
University of Southern California (04/10/08) Guerrero, Jean
Project:Possibility, a nonprofit organization created at the University of
Southern California last year, is working to develop free and accessible
software for the disabled community. Project:Possibility chair Christopher
Leung, a USC computer science graduate, says he started the organization
because he wanted to use his engineering skills for a worthy cause and
encourage others to do the same. Leung says Project:Possibility will
develop programs to help disabled people complete everyday tasks, such as
helping the visually impaired dial calls on a cell phone or type words on a
keyboard. Last November, 30 students met for the project's first event, a
competition called "SS12, Code for a Cause," in which they were given 24
hours to create useful software for the disabled community.
Project:Possibility is now giving students 12 weeks to create whatever is
feasible that will empower users in some way. The 25 student developers
have been divided into five groups, each under the direction of a
representative from Google, Amgen, or NASA. Software under development
includes a search engine specifically designed to help people find
technology for the disabled, a cell phone currency reader for people with
visual impairments, a gesture recognition program that will translate
movement into computer actions, a Web captioning system, and word
prediction software that predicts the words people are trying to type based
on eye movements directed toward an on-screen keyboard.
Click Here to View Full Article
to the top
Cryptographers Speak of Threats, Voting, and Blu-Ray
Rumors
CNet (04/08/08) Vamosi, Robert
The creators of the Diffie-Hellman key exchange and the EMC security
division discussed the state of security over the past year and answered
questions posed by a moderator during the annual cryptographers' panel at
RSA 2008 in San Francisco. Sun Microsystems' Whitfield Diffie said pure
defense does not work on the Internet, and added that the government might
consider going after opponents where they live and using different ways to
shut them down. Stanford University professor Martin Hellman said thinking
something is 99.9 percent safe is the greatest risk, and he warned about
becoming complacent. MIT professor Ronald Rivest, who is part of a group
that has released a public proposal on voting system standards, said he
favors software-independent voting systems, which do not entirely depend on
software and use paper or another way to capture votes. Meanwhile,
Weizmann Institute of Science professor Adi Shamir said the rumor that
Blu-Ray offers better overall security than HD DVD could be a sign that
security is becoming a factor in consumer electronics.
Click Here to View Full Article
to the top
Professor Studies Use of Robot Teams for Rescue,
Soccer
Daily Texan (04/10/08) Winchester, Lauren
After developing a team of soccer playing robots, University of Texas (UT)
at Austin professor Peter Stone will turn his attention to collaborative
rescue missions. Stone recently received a 2008 Guggenheim Fellowship, and
he plans to spend a year in Israel to study the use of teams of robots for
rescue situations following natural disasters. "At a disaster scene,
people go and look for victims, but it is less risky for the building to
collapse on robots," says Stone, in reference to the Sept. 11, 2001,
terrorist attacks. "People brought in individual robots to help with the
rescue." The reason why robots are often unable to coordinate and share
information with each other is because they are programmed by different
people. Stone started the UT Austin Villa robotic soccer team in 2002, and
this effort would inspire his research into the interaction of robots
programmed by different individuals. The UT Austin soccer robots were
programmed by the same group, but Stone had to figure out a way to get the
robots to play against other teams that were programmed differently.
Click Here to View Full Article
to the top
Open Source 3D Printer Copies Itself
Computerworld New Zealand (04/08/08) Hedquist, Ulrika
The Replicating Rapid-prototyper printer (RepRap) is an open source,
self-copying 3D printer that works by building objects in layers of
plastic, primarily polylactic acid, a bio-degradable polymer made from
lactic acid. Unlike existing prototyping printers, RepRap can replicate
and update itself, including printing its own parts, says RepRap software
developer Vik Olliver. The RepRap development team, is spread throughout
New Zealand, the United Kingdom, and the United States. By making the
project open source, the team hopes to be able to continue to improve the
machine until it can do what people want it to do. Improvements received
by the team are then sent back to users, allowing RepRap to evolve as a
whole. A recent feature added to RepRap are heads that can be changed for
different kinds of plastic. Olliver says a head that deposits low
melting-point metal is in development, which means low melting-point metal
could be put inside higher melting-point plastic, allowing for the
production of structures such as motors. RepRap could also allow people to
build circuits in 3D and in various shapes. Having the machine be able to
copy itself is the most useful feature the team can give it and is the
primary goal of the project, Olliver says.
Click Here to View Full Article
to the top
Memory in Artificial Atoms
University of Copenhagen (04/07/08)
Nano-physicists Jonas Hauptmann, Jens Paaske, and Poul Erik Lindelof with
the Nano-Science Center and the Niels Bohr Institute at the University of
Copenhagen have demonstrated that when using carbon nanotubes as
transistors a single electron spin can be controlled. "One can picture
this single electron spin caught on the nanotube as an artificial atom,"
Hauptmann says. For several years, researchers have considered direct
electrical control over a single electron spin to be a theoretical
possibility, but the Copenhagen team has demonstrated the mechanism in
practice for the first time. Their new transistor concept makes use of a
carbon nanotube or a single organic molecule, rather than the traditional
semiconductor transistor. "Our discovery shows that the new transistor can
function as a magnetic memory," Paaske says, which will allow much faster
and more accurate computer data storage in the future.
Click Here to View Full Article
to the top
EC to Encourage 'Internet of Things' Standards
Using RFID (04/07/08)
Global standards, regulatory, and other issues concerning radio frequency
identification and its role in making an "Internet of Things" a reality
will be the focus of a strategic EU-funded 7th Framework effort that will
include the participation of a group of international partners representing
Europe, the United States, Japan, Korea, and China. Coordination and
Support Action for Global RFID-related Activities and Standardization
(CASAGRAS) says the construction of a "global integrated intelligent
infrastructure that will exploit developments in pervasive networking and
interfacing with the physical world through existing and future
developments in RFID and associated technologies" is the goal of the
initiative, and it is the group's intention to supply a framework of
foundation studies to help the European Commission and the global community
define and address international issues and developments relating to RFID,
with a specific concentration on the Internet of Things. The CASAGRAS
partners will engage in a review of standards and procedures for
international standardization tied in with RFID, regulatory issues and
global numbering systems in relation to RFID standards, RFID utilization's
socioeconomic elements, application and standards areas, functional
developments and associated standards, and RFID's implications for
ubiquitous computing, networking, and the Internet of Things. The RFID
Global networking facility will be established in alliance with the new
European Center for automatic identification and data capture for the
purpose of enabling global RFID stakeholders' participation and
contribution to program development.
Click Here to View Full Article
- Web Link May Require Free Registration
to the top
Standardizing the Brain-Machine Interface
IEEE Spectrum (04/08) Vol. 45, No. 4, P. 16; Peck, Morgen E.
Operating a neural prosthetic requires algorithms to decode brain signals
in order to drive whatever device the subject is attempting to move or
manipulate by thought, and some people in the field of brain-machine
interface research say the time has come to develop a standard decoder
algorithm. One project in this vein is an effort led by MIT computer
scientist Lakshminarayan Srinivasan, whose goal is to unify elements from
all algorithms designed at major brain-machine interface labs in an attempt
to create a generic approach that supports and augments each design.
Approaches to neural prosthetics have varied across different projects,
with some interfaces featuring direct brain or skull implants and others
featuring electrodes attached to the scalp. Srinivasan's algorithms have
performed as well or better than those he sought to bring together, at
least in simulation. But simulation is no measure for evaluating an
algorithm, says Duke University Medical Center engineer Mikhail Lebedev.
Brain-machine interfaces involve the adaptation of the brain and the
algorithms to one another. The brain also learns, to a certain degree, how
to game the algorithms' rules to achieve the desired outcome, so the
algorithms' performance cannot be completely predicted. Srinivasan says he
is currently learning electrophysiology methods and will soon begin human
tests of his algorithms.
Click Here to View Full Article
to the top