Association for Computing Machinery
Welcome to the November 21, 2011 edition of ACM TechNews, providing timely information for IT professionals three times a week.

ACM TechNews mobile apps are available for Android phones and tablets (click here) and for iPhones (click here) and iPads (click here).


Quietly, Google Puts History Online
New York Times (11/20/11) Eric Pfanner

The Google Cultural Institute is digitizing historical artifacts such as the Dead Sea Scrolls and making them accessible to any Internet user. "We’re building services and tools that help people get culture online, help people preserve it online, promote it online, and eventually even create it online," says Google Culture Institute director Steve Crossan. Google also recently helped put photos and documents from the Yad Vashem Holocaust Memorial on the Internet. The institute currently is partnering with the Nelson Mandela Foundation in South Africa and the Palace of Versailles to help it develop galleries dedicated to the history of the chateau. Crossan says that Google Cultural Institute researchers also are developing a standard set of tools that institutions can use to digitize their collections. "We hope we bring competence in storing large amounts of data and serving it and creating a good experience for users, but we’re not professional curators or historians or artists ourselves, so we need to connect with that world," he says.

Mining the Language of Science
University of Cambridge (11/18/11)

University of Cambridge researchers are developing computer systems that can understand written language the same way that humans do. One of the projects, called CRAB, is a text-mining system for assessing the cancer risk associated with certain chemicals, says Cambridge's Anna Korhonen. Literature review is the first step of risk assessment, but also is a major bottleneck, as there could be thousands of articles for a single chemical, Korhonen notes. CRAB's text-mining technology features software that can analyze natural language texts, regardless of their complexity, consistency, or ambiguity. CRAB takes the textual content of each relevant U.S. National Library of Medicine biomedical bibliographic database abstract and classifies it according to taxonomy. Korhonen says that CRAB can quickly build a profile for any particular chemical based on all of the available literature. "In a recent experiment, we studied a group of chemicals with unknown mode of action and used the CRAB tool to suggest a new hypothesis that might explain their male-specific carcinogenicity in the pancreas," she says.

Google's Search Algorithm Challenged
IDG News Service (11/19/11) Philip Willan

Padua University professor Massimo Marchiori is leading the development of Volunia, a new search engine that could challenge Google's search algorithm and lead to radically different search engines in the future. "It's not just Google plus 10 percent. It's a different perspective," says Marchiori, who contributed to the development of Google's search algorithm. "It's a new radical view of what a search engine of the future could be." Volunia's Web site allows visitors to sign up for a chance to test the beta version of the search engine, which will be launched in 12 languages by the end of the year. "If I didn't think it was something big, capable of competing with the giants of online search, I would never have got involved," Marchiori says. The project is headquartered in Padua, with funding being supplied by Sardinian entrepreneur Mariano Pireddu. "The difference of our search engine is what will enable us to emerge," Marchiori says. Pireddu says the Volunia researchers are not attempting to build a better search engine than Google's, but rather they are trying to create a different kind of search engine that can work alongside Google's.

New Stanford Software Takes [email protected]'s Biological Research to Supercomputers
Stanford Report (CA) (11/18/11) Sarah Jane Keller

Stanford University researchers have developed Copernicus, a distributed framework for supercomputers that is based on [email protected], a distributed computing project also developed at Stanford that borrows computing time from home computers to simulate how proteins take shape. "We're bringing [email protected] to supercomputers," says Stanford researcher Vijay Pande. Copernicus enables other researchers to run simulations, including molecular models, using processor time on multiple supercomputers or computing clusters, instead of on home computers. "It opens the door to huge crowds of people using these methods, which have matured with [email protected]," Pande says. Copernicus uses the fast communication available between supercomputers, combined with statistical sampling methods, to run parallel simulations within or between supercomputers, in a process known as strong scaling. "This method should be able to use any supercomputer on the planet completely," Pande says.

Obfuscated Code Contest Returns
IDG News Service (11/16/11) Joab Jackson

The International Obfuscated C Code Contest (IOCCC) recently launched a challenge for the first time in five years, asking participants to write bizarre and unnecessarily complex C programming code. The goal of the contest is to stress to programmer the importance of good programming style. The contest also has some secondary benefits, such as exposing both structural quirks of the C language and weaknesses of the standard compilers. Winning programs have "pushed the boundaries of the C language," says IOCCC co-founder Landon Curt Noll. The judges are looking for code written in such a way that another programmer would have extreme difficulty determining what the program actually does, according to Noll. Adding to the challenge is the fact that the rules are not entirely straightforward, beyond a few basic principles, such as the code must be able to be compiled on a standard ANSI C compiler and must come with documentation that clearly states what the program is supposed to do. IOCCC will accept submissions for this year's contest beginning Dec. 1, and the deadline for submissions is Jan. 12, 2012. "One of the things that the contest emphasizes is that having a working program is not enough," Noll says. "The code has to be maintainable and adaptable."

App Boosts the Sounds You Have Trouble Hearing
Technology Review (11/17/11) Kenrick Vezina

Smartphone users who have trouble hearing could benefit from ACEHearing, an application designed to test for their specific form of hearing loss and then customize the audio output of their device. ACEHearing enables the diagnostic test via headphones, by having a smartphone play sounds across a range of frequencies and asking whether the user can hear them. The software adjusts the audio output by amplifying the frequencies that cause the most trouble for the user, rather than simply turning up the volume on a device. The test is equivalent to those administered by traditional audiologists, according to the results of clinical trials. Chinese University of Hong Kong's Andrew van Hasselt, one of the app's principal developers, says ACEHearing is a proof-of-concept prototype that is part of a larger effort to add the functionality to any device that produces sound. The Hearing Loss Association of America's Barbara Kelly notes that it can take people up to seven years to address a hearing loss, and ACEHearing's researchers say the app is designed to accelerate that process by providing an easy-to-use solution.

Smart Listeners and Smooth Talkers
University of Cambridge (11/17/11)

Researchers at the universities of Cambridge, Edinburgh, and Sheffield are working to lay the foundation for a new generation of speech technology to interact naturally with people, and display human-like performance and behavior. New systems would need to automatically adjust to different speakers, efficiently detect who said what, when, and how, and learn from mistakes. "Our approach is to build systems that are trained on a very wide range of data types and enable detailed system adaptation to the particular situation of interest," says Cambridge professor Phil Woodland. "To access and structure the data, without needing manual transcripts, we are developing approaches that allow the system to train itself from a large quantity of unlabelled speech data." The technology would need to account for the expressiveness of speakers and their intention in speech to respond to an individual's voice, vocabulary, accent, and expression. The researchers plan to develop a personalized voice-controlled device to help the elderly interact with control systems in the home, and a portable device that enables users to create a searchable text version of any audio they encounter on a daily basis, among other applications.

$50,000 to Solve the Most Complicated Puzzle Ever Attempted
UCSD News (CA) (11/16/11) Tiffany Fox

University of California, San Diego (UCSD) researchers are participating in the U.S. Defense Advanced Research Projects Agency's (DARPA's) Shredder Challenge, which involves piecing together about 10,000 pieces of different documents that have been shredded. The UCSD team will use a combination of crowdsourcing and computer-vision algorithms to solve the puzzle. The researchers will analyze crowdsourced partial solutions with algorithms for clustering pieces and finding which pieces are likely to fit together, which will make subsequent puzzle-solving easier for participants. The team devoted a great deal of time discussing how to create an incentive structure that will encourage participation, focusing on a referral-based crowdsourcing system that helped the Massachusetts Institute of Technology team win DARPA's Red Balloon Challenge in 2009. If UCSD wins, the amount of money a participant receives is based on how many edges they connected in the puzzles. For every edge connected by the participant, he or she would receive $1. The person who recruited that participant would receive 50 cents, and the person above that recruiter would receive 25 cents. "Recruiting players is essential to solving all the puzzles, especially as they get bigger," says UCSD’s Manuel Cebrian.

A Touchscreen You Can Really Feel
Ecole Polytechnique Federale de Lausanne (11/16/11) Sarah Perrin

The texture of a touchscreen can be controlled to provide users with the feeling of modifying specific areas with their fingertips, according to researchers at the Ecole Polytechnique Federale de Lausanne Integrated Actuators Laboratory (LAI). The team has developed a new generation of tactile surfaces with relief effects, and is designing the technology for smartphones, tablets, computers, and vending machines. "We're adding the sense of touch to tactile surfaces," says LAI Ph.D. student Christophe Winter. "The term 'touchscreen' that's used to describe current technology is really a misnomer, because they only provide visual and auditory feedback." The team uses piezoelectric materials, which vibrate when a voltage is applied to them, to create a tactile surface. The researchers say that enabling users to feel actual raised keys under their fingers could improve the usability of devices and improve the user experience. The technology also could make devices more accessible for the visually impaired, make video games more entertaining, and enrich online texts.

Squishybots: Soft, Bendy and Smarter Than Ever
New Scientist (11/16/11) Justin Mullins

A squishy, tentacled configuration may be an accurate design model for future robots, as a rigid humanoid shape is proving impractical for many of the tasks people want robots to perform. A key element of such designs is morphological computing, a discipline that holds that a robot's intelligence can be enhanced through the optimization of its body's interaction with its environment. This represents an inversion of conventional thinking, which dictates that an organism has a central processing capability where intelligence in housed, and its body's interaction with its surroundings demonstrates that intelligence. Using the embodied intelligence approach, researchers in Pisa, Italy, are building a soft, rubbery robot octopus equipped with appendages whose grasping ability exceeds that of the most advanced robots. Another speculative application of morphological computing principles is a soft robot surgeon concept that Kings College London researchers are studying. The robot would enter the body through a natural orifice or incision, pass soft tissues and organs without impediment, and harden once in place. The advantage of the embodied intelligence strategy is that the robots will be ideally suited for the job at hand.
View Full Article - May Require Free Registration | Return to Headlines

Rankings Released for Supercomputers Doing "Big Data"
Sandia National Laboratories (11/15/11) Neal Singer

This year's Graph500 competition, which measures supercomputers that handle big data scaling problems, featured 50 systems, up from nine in last year's competition. The U.S. National Nuclear Security Administration/SC Blue Gene/Q Prototype II ranked first, while Sandia National Laboratories' Ultraviolet, Red Sky, Dingus, and Wingus ranked 10th, 13th, 23rd, and 24th, respectively. The Graph500 focuses on a machine's ability to solve complex problems involving random-appearing graphs. "Companies are interested in doing well on the Graph500 because large-scale data analytics are an increasingly important problem area and could eclipse traditional high-performance computing in overall importance to society," says Sandia's Richard Murphy. Big-data problems indicate how well computing systems store and communicate large amounts of data in irregular, fast-changing communication patterns. Moscow State University's Lomonosov supercomputer ranked second on this year's Graph500 list, followed by the Tokyo Institute of Technology's TSUBAME, Forschungszentrum Julich's Jugene, and Argonne National Laboratory's Intrepid.

NC State Team to Develop Energy Efficient 3-D CPU
NCSU News (11/15/11) Matt Shipman

A $1.5 million grant from Intel will enable a North Carolina State University (NCSU) team to solve some problems involving the development of three-dimensional (3D) central processing units (CPUs). The researchers will build a 3D CPU chip stack, and their improvements to architecture and circuits could boost performance per unit of power by 15 percent to 25 percent. The team wants to address how to reconcile chips that are designed and manufactured in different places to different specifications so that they can work together in 3D, tackle questions concerning heat dissipation, and focus on test and yield challenges such as how manufacturers can test individual CPU components to ensure they are functional. The target date for the development of a complete prototype is 2014. Three-dimensional integrated chips are vertically integrated chips that are connected by vertical electronic connections--called through silicon vias--that pass through silicon wafers. Conventional computer chips only operate in two dimensions. NCSU professors Paul Franzon, Eric Rotenberg, and Rhett Davis will collaborate with Duke University's Krishnendu Chakrabarty on the research.

Abstract News © Copyright 2011 INFORMATION, INC.
Powered by Information, Inc.

To submit feedback about ACM TechNews, contact: [email protected]
Current ACM Members: Unsubscribe/Change your email subscription by logging in at myACM.
Non-Members: Unsubscribe