Association for Computing Machinery
Welcome to the December 4, 2013 edition of ACM TechNews, providing timely information for IT professionals three times a week.

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


Caregiving Goes Digital--and Lets Boomers Age in Place
USA Today (12/02/13) Marco Della Cava

Smart homes are gaining momentum as a way to help care for elderly people remotely by using sensors to monitor activity and offer reminders to take medications. Companies are developing and releasing products such as bottle caps that glow when it's time to take medicine, chairs that read vital signs, and carpets that analyze walking patterns to predict deterioration and potential falls. The technology is likely to benefit from wealthy, aging Baby Boomers who wish to remain in their homes, even as the number of caregivers drops. In 2010, 80-year-olds had an average of 7.2 middle-age caregivers, but that number is expected to drop to 2.9 by 2050, according to an American Association of Retired Persons report. "With fewer people to watch over tomorrow's seniors, some will move into digitally connected retirement homes, while others will simply retrofit their own homes," says Institute for the Future research director Jason Tester. "The home will serve as an early detection system. Research shows that even small changes in daily habits can hint at serious problems to come." However, the technology faces several obstacles to widespread adoption, including interoperability of products from different companies, privacy issues due to the sensitive data being collected, and an aversion to monitoring among some people.

Japanese Robots Earn Their Keep
Financial Times (12/02/13) Jennifer Thompson

Japanese researchers have developed Kirobo, a robot designed to provide assistance and companionship to human astronauts who spend months working in space. Kirobo is programmed to communicate in Japanese and recognize voices and faces. It also is capable of holding a conversation by improvising basic responses. Although Japanese roboticists have come under scrutiny for making robots that entertain rather than serve practical purposes, Kirobo's developers think it can do both. "I believe robots will be the next smartphone just like Google believes Google Glass will be the next smartphone," says Kirobo's creator Tomotaka Takahashi, a robotics engineer at Tokyo University. Kirobo recently was sent to the International Space Station in an unmanned rocket. "Interest in space robots is growing more quickly now than 10 years ago," notes the U.S. National Aeronautics and Space Administration's Ron Diftler, manager of the Robonaut project, which seeks to build the first humanoid robot that can help humans work and explore in space. "The space station is a very busy place and having an extra set of hands, in this case robot hands that can handle maintenance tasks, frees up the crew for more science."
View Full Article - May Require Free Registration | Return to Headlines | Share Facebook  LinkedIn  Twitter 

This Landmark Study Could Reveal How the Web Discriminates Against You
Forbes (12/02/13) Parmy Olson

Researchers at Princeton University and KU Leuven have released "bots" that imitate humans in order to monitor how they experience websites differently. The simulated profiles feature varying ages, genders, locations, incomes, interests, and other factors to enable the identification of patterns of discrimination. For example, some websites display different prices and deals based on where their shoppers reside, but firms that track such activities tend to operate covertly. The researchers will use the bots to analyze ads, prices, offers, and emails received by the simulated profiles over the coming months to identify discriminatory patterns across different sites. The study will be led by Princeton professor Arvind Narayanan, who plans to use his team's findings to form a "Web privacy census" that maps which companies are collecting which private data, who they are sharing with, and what conclusions they are drawing from it. Narayanan says the researchers will use a tracking process called automated reverse engineering, which will enable them to detect attributes such as behavioral segments and parameters that their bots may fall into during ad auctions. He says the process, which involves "a single, extensible infrastructure with various plugins" to measure personalization or discrimination, should provide more information than previous research based on manual or crowdsourcing techniques.

Study: Black Women Falling Behind in STEM Fields
New Pittsburgh Courier (12/01/13) Zenitha Prince

Women of color, especially black women, are significantly underrepresented in academic science, technology, engineering, and mathematics (STEM) positions, even as the U.S. focus on STEM education intensifies, according to a study by the Institute for Women's Policy Research. In 2010, minority women professors accounted for 2.1 percent of STEM faculty at four-year colleges and universities in the United States, although they represent 13 percent of the U.S. working-age population. The study also found that 6,400 women of color with STEM doctorates hold assistant, associate, or full professorships, compared with 19,800 white women, 20,500 men of color, and 65,100 white men. The life sciences field employed the largest number of underrepresented minority women faculty, while computer science and mathematics employed the least. The report cites hostile work environments, insufficient mentorship, difficulty with work-life balance, and the lack of a multicultural perspective in many academic departments as factors that limit women of color in academic STEM careers. To improve the situation, the report recommends providing targeted funding to women of color, changing hiring and promotion policies, offering culturally sensitive mentors, and creating a scorecard system to gauge the diversity progress of institutions.

New Algorithm Finds You, Even in Untagged Photos
University of Toronto (12/02/13)

University of Toronto researchers have developed an algorithm that can change the way photos are found on social media websites. The search tool uses tag locations to quantify relationships between people, even those who are not tagged in a specific photo. The researchers say their algorithm, called relational social image search, achieves high reliability without using object- or facial-recognition technology. "Our algorithm is simply based on the number of tags, not on the number of photos, which makes it more efficient to search than standard approaches," says Toronto professor Parham Aarabi. He notes that although the algorithm's interface currently is used just for research, it could be incorporated on the back-end of large image databases and social networks. "I envision the interface would be exactly like you use Facebook search--for users, nothing would change," Aarabi says. The researchers have also used the algorithm to generate maps; they tagged a few photos of buildings on the University of Toronto campus and ran them through the system with several untagged campus photos. "The result we got was of almost a pseudo-map of the campus from all these photos we had taken, which was very interesting," Aarabi says.

Magnetic Tape to the Rescue
The Economist (11/30/13)

Magnetic tape is being revived as a viable storage medium for the vast volume of data generated, which doubles every two years at its current rate. The European Organization for Nuclear Research's (CERN) Alberto Pace says magnetic tape has several advantages over hard disks for long-term data preservation, including speed, reliability, zero power consumption for storage, and security. For example, a broken tape can be spliced back together and only lose a few hundred megabytes of information, while deleting 50 petabytes of CERN data on magnetic tape would take years, rather than minutes for disk-based data. IBM Zurich research lab's Evangelos Eleftheriou cites tape's lower cost compared to disks, and much greater longevity of the stored data. However, the looming flood of data is too much even for modern tape cartridges to handle, and higher densities are required. Eleftheriou currently is developing a tape with a density of 100GB per square inch, as well as creating the equipment needed to read it. The technology could potentially yield a cartridge capable of storing more than 100 terabytes, and a key challenge to be met is to position the read/write head to within 10 nanometers.

Wave Fingers, Make Faces: The Future of Computing at Intel
CNet (11/29/13) Shara Tibken

Intel is developing perceptual computing technology that will sense a user's emotions and body language. Perceptual computing uses gestures, facial recognition, and voice recognition to make devices more "natural, intuitive, and immersive," says Intel's Anil Nanduri. He says Intel aims to obtain "sensory inputs that make [computers] more human like." Cameras, for example, will enable devices to sense emotion and detect a person's biometric data. Computers also will be able to hold conversations and comprehend general, rather than specific, commands, taking context into account. Intel intends to teach devices to recognize depth using three-dimensional (3D) cameras, and has teamed with Creative to integrate 3D cameras into devices such as PCs and tablets in the second half of 2014. To spur developer interest in perceptual computing, Intel last year released a software development kit, which has been downloaded more than 26,000 times, and has sponsored perceptual computing contests. Intel believes it has advanced the ecosystem sufficiently to begin focusing on hardware, and hopes that devices with perceptual computing features will emerge next year. Although sensory inputs such as gestures will not be ideal for all uses, Intel intends to focus on applications that will prove genuinely useful. Gaming and education are two areas in which perceptual computing could gain momentum, Nanduri says.

Call-Log App Aims to Reverse-Engineer NSA Surveillance
Technology Review (11/28/13) Tom Simonite

Stanford University researchers have launched the MetaPhone Project, which asks volunteers to install an Android app that sends the researchers copies of their smartphone's call logs and basic data from their Facebook account. The researchers believe that a large collection of such data will make it possible to use data-mining techniques to discover which aspects of people lives can be revealed by examining just their calling and texting logs. The project is meant to simulate the U.S. National Security Agency's (NSA) controversial data collection programs. "Some defenders of the NSA’s bulk collection programs have taken the position that metadata is not revealing," says MetaPhone Project co-founder Jonathan Mayer. "We want to provide empirical evidence on the issue." Using the small amount of data already collected by the app, the researchers have shown that calling and texting patterns can reveal whether a person is in a relationship. They say their findings could have influence beyond policy and legal debates about domestic surveillance.

Patients Drive New Wheelchair With Their Tongues
Christian Science Monitor (11/27/13) Elizabeth Barber

Georgia Tech researchers have developed the Tongue Drive System (TDS), a wearable system that enables paralyzed people to control wheelchairs with just flicks of their pierced tongues. The researchers say TDS could help patients disabled from the neck down access their worlds much more easily than they can with current assistive systems. The TDS system involves a magnetic tongue stud that relays the wearer's tongue movements to a headset, which sends the commands to a smartphone or other Wi-Fi connected device. "This is one assistive device for your entire daily life," says Georgia Tech professor Maysam Ghovanloo. The researchers are hoping to develop an even smaller version of the device, called iTDS, that could disappear entirely into a patient's mouth and "eliminate the risk of stigma" of having a tongue stud, Ghovanloo notes. During testing, the researchers compared TDS to another popular assistive system known as sip-and-puff. In one trial, the participants were timed driving their wheelchairs around a series of obstacles and making certain maneuvers, using both TDS and sip-and-puff. The researchers found that TDS is, on average, three times faster than sip-and-puff, and just as accurate.

Argonne Lab Taking Next Steps to Exascale Computing
InformationWeek (11/26/13) Patience Wait

Scientists at the Argonne National Laboratory's Mathematics and Computer Science Division working on the Argo project are creating a prototype exascale operating system and runtime software that would reach the exaflop mark. Scientists from the Lawrence Livermore and Pacific Northwest National Laboratories and several universities are collaborating with the Argonne team on the project, which is funded through a $9.75-million Department of Energy Office of Science grant. Computer chips are no longer making performance gains, says Argo program manager Pete Beckman. "We've been turning up the clock every year, but we got to this point at about 3 gigahertz where it really hasn't gotten any faster," Beckman says. "Instead, now companies are making them more parallel." Massive parallel processing requires both hardware and software changes, which Argo will address by developing an open source prototype operating system that can run on various architectures. The researchers aim to have the first prototype systems by the end of the three-year project, but experts predict full-scale exaflop computing will not be feasible until 2018. The researchers say additional computing power will enable breakthroughs in the most challenging scientific problems, such as understanding the workings of subatomic particles.

Robotutor Marks the Homework of a Class of Thousands
New Scientist (11/25/13) Hal Hodson

Although the scale and reach of massive open online courses (MOOCs) is growing every year, a human teacher cannot guide, correct, and give feedback to thousands of students all working at the same time. To address this problem, Stanford University researchers have developed Codewebs, an artificially intelligent tutor for online students that can analyze and assess submitted code. The researchers say Codewebs also can give students fast, tailored feedback and guidance. Codewebs runs machine-learning algorithms on a database of code submissions from thousands of students in courses offered by Coursera. The software breaks down the students' attempts at coding into small pieces and indexes them, enabling the system to compare the submissions with the database and cluster them according to their similarity with one another. The index lets a human instructor pick one submission, write feedback, and then have it automatically sent out to all of the students who wrote similar solutions. "If a student hands in homework, not only does it say, 'Good job, you solved it like your peers,' but if it looks like the student is solving the problem in a way that is detrimental to their learning, we could give feedback to push them away from that," says Stanford's Chris Piech.

Opening Up the Accelerator Advantage
HPC Wire (11/26/13) Tiffany Trader

The U.S. National Science Foundation (NSF) has awarded a grant of nearly $2 million for a project that seeks to move supercomputing capabilities beyond the domain of elite scientists. Researchers at the Georgia Institute of Technology and the University of Southern California will create tools designed to help developers take advantage of hardware accelerators in a cost-effective and power-efficient manner. They will make use of tablets, smartphones, and other Internet-era devices, says Georgia Tech's David Bader. "We want to take science that used to be only available to elite scientists and bring that to everyone around the planet," Bader says. "We are bringing supercomputing to the masses." Over the next three years, the researchers will work on different types of optimizations for XScala, the software framework for developing efficient accelerator kernels. The project also will focus on security and social network analysis. In addition, the team will focus on XBazaar, a public software repository and forum that is similar to an app store. "XBazaar will serve as a one-stop shop for high-performance algorithms and software for multi-core and many-core processors," according to the NSF announcement.

Spectral Redemption: Finding the Hidden Groupings in Networks
Santa Fe Institute (11/27/13)

Researchers at the University of California, Berkeley, the Sante Fe Institute (SFI), and the French National Center for Scientific Research (CNRS) say they have developed an approach to cluster detection that addresses weaknesses that conventional techniques have in dealing with networks where most nodes have just a few links. Historically, researchers have struggled to find a spectral method that is computationally efficient, but that also finds groupings in networks down to the theoretical limit. The researchers' method, called the non-backtracking operator, specifies that during analysis, information flowing from node to node may not immediately return from its source. The researchers tested the non-backtracking technique on several network datasets commonly used to benchmark clustering methods, including several real-world networks, and found that it is reliable to the theoretical limit. "This work shows how crucial it is to build connections between scientific communities," says Berkeley's Elchanan Mossel. "Bringing together concepts, methods, and points of view from statistical physics [SFI and CNRS] and mathematics [Berkeley] gave us a whole that is much greater than the sum of the parts."

Abstract News © Copyright 2013 INFORMATION, INC.
Powered by Information, Inc.

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