Association for Computing Machinery
Welcome to the November 9, 2012 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).

HEADLINES AT A GLANCE


The Internet Remains a Tangle of Issues
New York Times (11/08/12) Somini Sengupta; Edward Wyatt

Among the Internet-related policy issues President Barack Obama is likely to face in his second term is congressional polarization over regulations limiting the use of personal information online, statutes for protecting critical infrastructure from cyberattacks, policies to contend with online piracy, and pressure from the technology industry to implement tax code reforms and more visas for foreigners with much sought-after math and science skills. The administration will probably continue its push for freeing up wireless spectrum to fulfill the needs of smartphones and tablets to send a variety of information. Obama is expected to resume his campaign for the enactment of cybersecurity legislation to protect critical infrastructure, and the primary hurdle is whether the government should be permitted to regulate how private infrastructure operators shield their systems from cyberattacks. Members of Congress also are expected to lobby for several privacy-related bills, one involving updates to the Electronic Communications Privacy Act covering how government authorities can wiretap cell phone and Internet communications. The Internet industry also wants a 1988 law protecting the privacy of movie rentals eased, and the White House could once again find itself in the middle of a battle between the industry and Hollywood over online piracy.


Microsoft Brings Star Trek's Voice Translator to Life
Technology Review (11/08/12) Tom Simonite

Microsoft has developed and publicly demonstrated nearly instantaneous Chinese-to-English spoken word translation software capable of preserving the distinctive cadence of the speaker's voice. The system illustrates the advancement of Microsoft's speech-recognition technology, which is founded on learning software based on the operating mechanism of neural networks. The system recognizes the speaker's words, rapidly transforming the text into properly ordered Chinese sentences that are then passed to speech synthesis software trained to reproduce the speaker's voice. The software performs voice synthesis by modifying a stock text-to-speech model so that it produces certain sounds in the same manner the speaker does. "Rather than having one word in four or five incorrect, now the error rate is one word in seven or eight," says Microsoft's Rick Rashid. Although he concedes the system is far from infallible, it offers sufficient capability to enable communication where none would otherwise be possible. "We don't yet know the limits on accuracy of this technology--it is really too new," Rashid notes. "As we continue to 'train' the system with more data, it appears to do better and better."


Cambridge Software Improves Quality of Sound for Hearing Aid Users
University of Cambridge (11/08/12) Sarah Collins

Cambridge University researchers have developed CAM2, software that improves the quality of sound for hearing aid users. The software dictates the amount of amplification of high-frequency sounds required to restore their audibility. The technique increases the frequency range of sound that individuals with hearing loss are able to detect, improving speech perception, sound localization, and the ability to hear certain musical sounds. "Until recently, hearing aids only provided amplification for frequencies up to four or five kHz, whereas a person with normal hearing can hear for frequencies up to 15 or 20 kHz," says Cambridge professor Brian Moore. However, he says CAM2 extends the fitting range up to 10 kHz. The researchers note a recent study compared the CAM2 method with NAL-NL2, which is a fitting method used by the U.K.'s National Health Service and many other health organizations around the world. A majority of participants in the study preferred CAM2, both for overall sound quality and for the clarity of speech in a noisy situation.


Counting Votes, in the Precinct and on the Web
Harvard University (11/06/12) Caroline Perry

Harvard University researchers have found that all multi-agent systems, including national elections, can be improved using theories and techniques in computer science. For example, without the language to express conditional preferences, the whole election system can suffer, says Harvard's Lirong Xia. "In an online context or in artificial intelligence, if you're talking about a thousand alternatives, the voters won't even know their full preferences," Xia says. He also notes that instead of voting for a preferred third-party candidate who has little chance of success, many voters will vote for their second choice to prevent their least-favorite candidate from winning. However, if voters were able to express their preferences in order, the winner would be more difficult to calculate but the outcome might be more satisfactory. Online voting systems could offer an opportunity to change the process, but they currently have security and privacy shortcomings. Xia and his colleagues are trying to develop a system that can determine the threshold at which vote manipulation is quantifiably difficult to achieve, which could help guarantee the security of online voting systems. "I’m trying to design a faster way to detect fraud with high statistical confidence," Xia says.


Inside the Secret World of the Data Crunchers Who Helped Obama Win
Time (11/07/12) Michael Scherer

President Barack Obama's successful re-election owes a lot to a massive backroom effort that used insights mined from data gathered, stored, and analyzed over the two-year campaign for the purpose of planning strategy. The initiative helped Obama raise $1 billion, reconfigured the TV ad-targeting process, and generated models of swing-state voters that could be used to boost the campaign's efficacy. The effort's first 18 months involved the consolidation of numerous databases into a megafile that could integrate information gleaned from pollsters, fundraisers, field workers, and consumer databases, as well as social-media and mobile contacts with the swing states' main Democratic voter files. The system enabled the data processors to run tests forecasting which kind of appeals would persuade specific kinds of people. Data collection and analysis also were vital to the success of an email fundraising campaign responsible for a large percentage of the money raised online. These methods were later repurposed to generate votes for Obama through the analytics team's utilization of several streams of polling data to construct a picture of voters in key states. The online get-out-the-vote campaign also employed Facebook on a massive scale to reproduce field organizers' canvassing activities.


Computational Text Analysis Made Possible Regardless of Language or Domain
Aalto University (11/08/12)

Aalto University's Mari-Sanna Paukkeri has devised computational methods for processing and analyzing online text regardless of its language or domain. The methods employ algorithms that sift through textual data sets for statistical dependencies and structures, from which specific text properties can be extracted. Paukkeri's area of concentration is how unsupervised machine learning applies to natural-language processing. Such techniques do not involve the manual pre-processing of the data set. Instead, the algorithms are left on their own to learn the nature of the data and what type of statistical dependencies and structures it contains. Paukkeri describes one method, Likey, that is applied to keyphrase and keyword extraction from text documents in 11 languages. Likey determines how common certain words and pairs, threes, and fours of words are in a data set. The keywords and phrases for a specific document are then defined, according to their frequency and context within the text. Paukkeri notes that methods where textual data is processed in all working languages can be especially beneficial for companies with a global reach.


In Bounties They Trust, but Does Paying for Security Bugs Make a Safer Web?
Wired News (11/08/12) Kim Zetter

Freelance security researchers who hunt for exploitable software bugs for cash rewards posted by vendors can sometimes make a decent wage, but some of the largest software vendors do not host bug bounty programs, which raises the question of whether such efforts have improved the security of the Web. Google says bug bounty programs have positively impacted security, as indicated in the decreasing numbers of incoming bug reports. Bug bounty programs also enhance security through the encouragement of responsible bug disclosure, as researchers inform vendors first so they can craft and issue a patch to customers before the information is publicized. "The mere fact that you have a bounty program shows you have a certain amount of [security] maturity, because it would be too expensive otherwise [to launch one]," says Veracode co-founder Chris Wysopal. "You could have your application reviewed by a third party for the price of just five bugs you might pay out [in a bounty program]." Although Google's organization-wide policy is to patch critical or serious bugs within 60 days of receiving a report, Facebook chief security officer Joe Sullivan believes this timeframe needs to be accelerated.


Twitter a Big Winner in 2012 Presidential Election
Computerworld (11/07/12) Sharon Gaudin

Both Twitter and Facebook played a big role in the 2012 U.S. presidential election, as this was the first national election since social-networking technology became mainstream. There was a lot of focus on Twitter during the primaries, and it was very widely used during the presidential debates. However Twitter use exploded on Election Day, as Twitter reported that about 31 million tweets were posted, peaking at 327,452 tweets per minute just after the national news networks began calling the election for President Obama. Facebook also was very actively used in the election. Countless users posted Facebook updates that they had voted and what the lines at the polling places were like. Meanwhile, Facebook posted a message on users' pages reminding them to vote and provided a link to the Facebook Polling Place Locator. ZK Research analyst Zeus Kerravala says social networking contributed greatly to the election, especially in terms of providing opinion broadly. "Twitter lit up the day before and the day of the election with people giving opinions on everything from legalizing pot to who to vote for," he says.


NASA, ESA Use Experimental Interplanetary Internet to Test Robot From International Space Station
NASA News (11/08/12) Rachel Kraft

Researchers at the U.S. National Aeronautics and Space Administration (NASA) and the European Space Agency have developed an experimental version of the interplanetary Internet to control an educational rover from the International Space Station. The researchers used NASA's Disruption Tolerant Networking (DTN) protocol, which could eventually be used to develop Internet-like communications with space vehicles and support habitats or infrastructure on other planets. The DTN architecture enables standardized communications similar to the Internet to function over long distances and through time delays associated with on-orbit or deep space spacecraft or robotic systems. As part of the experiment, space station Expedition 33 commander Sunita Williams used a NASA-developed laptop to remotely drive a small LEGO robot at the European Space Operations Center in Darmstadt, Germany. "The demonstration showed the feasibility of using a new communications infrastructure to send commands to a surface robot from an orbiting spacecraft and receive images and data back from the robot," says NASA's Badri Younes. The research is part of NASA's Space Communication and Navigation Program, which coordinates multiple space communications networks and network support functions to regulate, maintain, and grow NASA's space communications and navigation capabilities in support of the agency's space missions.


Mine Your Language: Software Decodes Company Reports
New Scientist (11/02/12) Douglas Heaven

National Chengchi University researchers have developed an algorithm to recognize meaningful phrases in company financial reports. The algorithm uses statistical models to automatically identify opinion patterns, which it defines as subjective phrases paired with an opinion holder. "Computer linguistics and automated textual-information processing are one of the new frontiers in the world of finance," says the University of British Columbia's Werner Antweiler. "This technique adds another tool to our statistical toolbox of text-mining algorithms." Although trading algorithms mostly rely on quantitative information, textual information should be considered as well, says National Chengchi University's Chao-Lin Liu. For example, he notes the software could highlight phrases that do not appear to tally with a company's stated earnings, prompting a financial analyst to take a closer look. "Numbers can be used to convey a picture that does not correspond to reality," says the CFA Institute's Vincent Papa. "The tone of a report is a very useful complementary piece of information."


U.Va. Computer Science Grad Student Develops ‘Musical Heart’
UVA Today (11/05/12) Fariss Samarrai

University of Virginia graduate student Shahriar Nirjon has developed Musical Heart, a biofeedback-based system that helps smartphones select music that will help get their users' heart pumping during exercise, or slow it down when they want to cool down and relax. Musical Heart uses ear phones equipped with a microphone that detects the pulse in the ear's arteries. The app selects songs that optimize the heart rate of users based on their activity level. An algorithm refines the music selection by storing heart rate data and calculating the effects of selected music on the rate. "We've designed Musical Heart to be convenient, non-invasive, personalized, and low cost," Nirjon says. The system learns to select music that will have a desired effect on heart rate customized to the individual user, based on the effects of past music selections on the heart rate. Musical Heart also can select music that helps the user relax. Nirjon currently is developing a generic acoustic-sensing platform for smartphones that hosts a suite of apps, one of which is Musical Heart.


Microsoft Sponsors a Hacking Contest--the Coding Kind
Network World (10/31/12) Tim Greene

Microsoft recently hosted a hackathon for 40 teams of coders to write applications that run on Windows 8-based devices. The program's goal is for the participants to learn to write the applications with the guidance of Microsoft mentors. The hackathon also is geared toward writing apps for Azure, Microsoft's cloud-based platform service, and participants also attended Microsoft's Build 2012 conference for developers. The hackathon's winners take home $2,500 for the best application in three categories. Microsoft mentors are tasked with helping the participants take advantage of Windows 8's new touch-interface features and generating apps that look good. For example, Vesa Vainio, a participant from Finland, is developing Share My Photos, an app that stores photos in Azure for access by Facebook members. Although Facebook supports that capability, Vainio says his app has a twist. "They can easily share photos with friends of their choosing and not give ownership of the photos to Facebook itself," Vainio says.


TACC Aids Cardiovascular Disease Research With Animation, Computational Efforts
Texas Advanced Computing Center (10/31/12) Faith Singer-Villalobos

Researchers at the University of Texas at Austin's Institute for Computational and Engineering Sciences (ICES), Texas Advanced Computing Center, and Faculty Innovation Center have developed a 14-minute animation to explain the underlying nature of vulnerable plaques and a potential clinical procedure for treatment with the goal of personalizing diagnostic and therapeutic interventions in patients. "Visualization is absolutely essential--there's no question about it," says ICES professor Thomas Hughes. The animation illustrates how engineering and medical approaches can come together to address an unsolved clinical need, says ICES researcher Shaolie Hossain. The researchers used more than 12 software programs, including magnetic resonance imaging viewing applications and computer-assisted design and animation packages, to develop a program that performs some of the file conversions. The researchers also developed a computational toolset and simulation capabilities to model important characteristics of the research, such as fluid flow, drug release, nanoparticle properties, and patient-specific geometries. "Modeling these very complicated systems involves solving millions of equations each time step, and for millions of time steps, so the computational burden is enormous," Hughes says.


Abstract News © Copyright 2012 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