Association for Computing Machinery
Welcome to the March 24, 2014 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).


Parallel Programming May Not Be So Daunting
MIT News (03/23/14) Larry Hardesty

To overcome the theoretical limitations of "lock-free" parallel algorithms, which are used by developers to write programs for multicore chips, computer scientists have demonstrated "wait-free" algorithms, which guarantee all cores will make progress in a fixed span of time. However, deriving sequential code from wait-free algorithms is very complicated. Now, in a paper to be presented at ACM's Annual Symposium on the Theory of Computing in May, researchers at the Massachusetts Institute of Technology (MIT), the Technion–Israel Institute of Technology, and Microsoft Research will demonstrate an analytics technique suggesting that, in a wide range of real-word scenarios, lock-free algorithms can give wait-free performance. "In practice, programmers program as if everything is wait-free," says MIT professor Nir Shavit. "What we are exposing in the paper is this little-talked-about intuition that programmers have about how [chip] schedulers work, that they are actually benevolent," Shavit says. The researchers say the chip's performance as a whole could be characterized more simply than the performance of individual cores because the allocation of different chunks of code executed in parallel is symmetric.

NASA 'Codeathon' Seeks Apps That Address Coastal Flooding
Computerworld (03/20/14) Zach Miners

The U.S. National Aeronautics and Space Administration's (NASA) third global codeathon offers four climate-themed challenges, including one focused on coastal flooding. Participants will use federal data to create simulations and other technology to help people and businesses understand their exposure to coastal-inundation hazards and other dangers. The applications could reveal the extent to which they might be affected by rising sea levels and coastal erosion in the future. "Solutions developed through this challenge could have many potential impacts," says NASA chief scientist Ellen Stofan. The event offers more than 40 new challenges related to other areas, including robotics, human spaceflight, and asteroids. However, NASA is targeting social issues as well, and half of the challenges are focused on Earth-based problems. The codeathon, scheduled for April 12-13, will be hosted at nearly 100 locations around the world.

Internet2 Teams Up With India's National Knowledge Network
Chronicle of Higher Education (03/19/14) Megan O'Neil

Internet2 expects to formalize a partnership with its Indian counterpart this month. Internet2 officials believe a collaboration with India's National Knowledge Network would be among its most robust relationships with foreign education and research networks. Among research projects with budgets exceeding $100,000, there is more than $500 million in work already underway, but some observers would like to see funding for U.S.-India research rise into the billions. As part of the arrangement, India will commit to invest in fiber-optic capacity to connect its $1.33 billion education-research network to international education-research networks. There also are plans to establish video services that will enable research teams working in the two countries to interact. Researchers will collaborate in areas such as agriculture, public health, and the arts. "The work we have done has led up to an opportunity to make this collaboration a gold standard in joint efforts by two national research networks in support of research, education, and economic and social development," says Internet2 president H. David Lambert.
View Full Article - May Require Paid Subscription | Return to Headlines | Share Facebook  LinkedIn  Twitter 

Search Gets Smarter With Identifiers
CORDIS News (03/19/14)

European researchers are developing Okkam srl, a system that provides a centralized repository of identifiers for people, organizations, and things. As part of the Okkam srl project, the researchers have created a Global Open Naming System, an index of unique entities such as people, organizations, and products that lets people share data. "We provide a very fast and effective way of discovering data about the same entities across a variety of sources," says Okkam srl researcher Paolo Bouquet. Using basic technologies such as quick response (QR) codes and near-field communication, Bouquet says Okkam makes it possible to tag actual objects with online, up-to-the-minute data. "Okkam's entity naming system allows you to share the same identifiers across different projects, different companies, different data sets," Bouquet says. "You can always build on top of what you have done in the past." The project is a precursor to the emerging Internet of Things. For example, the province of Trentino in Italy has equipped each of its bus stops with QR codes, which makes it possible for passengers to scan the code and get the latest information about travel delays or download timetables.

How Your Tweets Reveal Your Home Location
Technology Review (03/21/14)

An algorithm developed by IBM researchers exploits anyone's last 200 Twitter postings to reveal their home city location with nearly 70-percent accuracy. The researchers filtered the Twitter channel for tweets that were geotagged with any of the largest 100 U.S. cities between July and August 2011, until they had pinpointed 100 different users in each location. They then downloaded the last 200 tweets posted by each user, rejecting privately posted messages until more than 1.5 million geotagged postings from almost 10,000 people remained. This dataset was divided in two, with 90 percent of the tweets employed to train the algorithm while the remaining 10 percent were used for testing it. The algorithm's underlying concept is that tweets contain key details about the user's likely whereabouts, and the researchers say tweet distribution throughout the day is roughly consistent across the country, so a user's pattern of tweets can offer solid clues to the tweeter's time zone. Testing the algorithm with the remaining data demonstrated that it correctly predicted tweeters' home cities 68 percent of the time, their home state 70 percent of the time, and their time zone 80 percent of the time when excluding obvious travelers.

Details Emerging on Japan's Future Exascale System
HPC Wire (03/18/14) Nicole Hemsoth

Japanese researchers are developing the next generation of the K Computer, which is expected to be the country's first exascale system, with the goal of the system being fully operational by 2020. Basic development for the future system, known as postK, is swiftly moving on software, accelerator, processor, and scientific project-planning fronts, according to Japan's Office for Promotion of Computing Science/MEXT's Yoshio Kawaguchi. PostK researchers say the system could be used to develop safer cars, new drugs with mitigated or reduced side effects, and better prediction models for natural disasters. However, to accomplish all of this at a reasonable cost will take significant innovation. University of Tsukuba researcher Mitsushisa Sato is focused on optimal accelerators for massive heterogeneous systems, which has led to the creation of extreme SIMD architecture designed for compute-oriented applications. The architecture is designed to address molecular dynamics and N-body-type simulations as well as stencil apps, and will aim for high performance in the area of around 10 teraflops per chip. University of Tokyo researcher Yutaka Ishikawa notes the many similarities in terms of challenges and potential problems the system could solve between K and the exascale system of 2020.

Pocket Diagnosis
University of Cambridge (03/19/14)

University of Cambridge researchers have developed Colorimetrix, a smartphone application designed to make monitoring certain diseases much clearer and easier for patients and doctors. The researchers say the app accurately measures color-based tests for use in home, clinical, or remote settings, and can transmit medical data from patients to healthcare professionals. The tests, normally in the form of small strips, work by producing color change in a solution, where the intensity of the color that is produced determines the concentration of that solution. However, the tests can be difficult to read accurately, and false readings are very common. Colorimetrix uses the phone's camera and an algorithm to convert data from colorimetric tests into a numerical concentration value on the phone's screen within a few seconds. The result can be stored, sent to a healthcare professional, or directly analyzed by the phone for diagnosis. The app has been shown to accurately report glucose, protein, and pH concentrations from commercially-available urine test strips without requiring any external hardware. "This app has the potential to help in the fight against HIV, tuberculosis, and malaria in the developing world, bringing the concept of mobile healthcare to reality," says Cambridge Ph.D. student Ali Yetisen.

Experts Explain Why Big Data Is a Big Deal
UCSD News (CA) (03/20/14) John B.B. Freeman

Expert speakers participating in a recent seminar at the University of California, San Diego discussed the rapid growth of big data and how it is affecting people's daily lives. California Institute for Telecommunications and Information Technology director Larry Smarr pointed to a typical Google search on a smartphone, whose operation requires more computing power than all of the Apollo space missions put together. "Never in our history have we had a sustained period of this kind of exponential growth [in computer science]," Smarr said. "What we're talking about is something humanity has never tried to deal with before." The key theme of the seminar was speculation on the future changes that big data will usher in. Fellow speaker and San Diego Supercomputer Center director Michael Norman discussed the center's Gordon supercomputer, which is a repository that moves, houses, and analyzes data with vast volumes of flash-based memory. The research areas Gordon is used for include climatology, finance, food production, big industry, physics, biological science, and government. Norman says the three central functions of big data are the volume of data, the speed of information produced, and the variety of data that is readily available.

When the Flu Bug Bites the Big Apple, Twitter Posts Can Tell the Tale
Johns Hopkins University (03/18/14) Phil Schneiderman

Johns Hopkins University (JHU) researchers previously have shown that Twitter posts can reveal flu trends at the national level. In a new study, researchers from JHU and George Washington University say tweets also can predict flu trends at the local level. The researchers sifted through billions of tweets to identify flu infections and where flu patients were located, isolated flu patient tweets to New York City, and compared their findings to the flu cases compiled by the city's Department of Health and Mental Hygiene. The team's results accurately gauged the spread of flu in New York City for the 2012-2013 flu season. The researchers used software developed at JHU to scan tweets and distinguish those from people who are likely ill, as opposed to people posting fears of catching the flu. Many Twitter users list the cities where they live or use a global positioning system-equipped phone to tweet. "We found that we could do just as well in predicting flu trends in New York City as we did nationally," says JHU professor Mark Dredze. "That's critical because decisions about what to do during a flu epidemic are largely made at the local level." The researchers believe their technique could be used to study crime, political developments, and responses to natural disasters.

Despite Pwn2Own 2014 Hacks, Application Sandboxing Still Critical (03/18/14) Brandan Blevins

Bug hunters and researchers attending the recent CanSecWest security conference's Pwn2Own hacking contest were able to successfully demonstrate 35 exploits of some of the most popular browsers and software suites. Despite these successes, participants and organizers say the results of this year's contest show that popular software is increasingly more secure and harder to compromise. Brian Gorenc with HP's Security Research group says most of the successful exploits had to target multiple vulnerabilities in order to succeed. "As the mitigations get added in to technologies, it is becoming more difficult," he notes. "It takes a significant amount of time to develop that chain of exploits." Chief among these mitigation techniques is the use of application sandboxing. Most major browsers and software have implemented sandboxing, thus requiring attackers to create exploits simply to defeat or escape the sandbox before they can exploit the targeted software. However, Carnegie Mellon University's Will Dormann says sandboxing alone is not enough. He suggests security professionals apply several other controls, such as running the Firefox browser's NoScript add-on and Microsoft's Enhanced Mitigation Experience Toolkit (EMET). EMET, for example, was able to mitigate all of the zero-day exploits of Internet Explorer used at Pwn2Own in 2013.

A 'Babelfish' Could Be the Web's Next Big Thing, Says AI Expert
The Guardian (United States) (03/14/14) Charles Arthur

University of Southampton professor Nigel Shadbolt thinks automatic real-time machine translation could be possible within 25 years, and will be just one of many potential innovations stemming from the evolution of artificial-intelligence systems facilitated by cloud computing. “With more cloud computing you can imagine that there will be more machine translation,” Shadbolt says. He says an automatic machine translation system will be achievable "because of the large-scale resources that will be available to do entire voice translation on the fly. I would be surprised if in 25 years we haven't got enhanced Bluetooth-based translators that you can just put in your ear." Dealing with real-time speech is currently a formidable challenge for machine translation and transcription systems, and thus far no machine has proven capable of conducting real-time transcription of conversations, although cloud computing could bring that goal close to fulfillment within several years. Two years ago, Microsoft showcased a system capable of performing almost real-time translation between English and Chinese with a lower error rate than previously demonstrated. "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.

Navigating User-Generated Resources: A Q&A With Computer Scientist Brent Hecht
Library Journal (03/18/14) Matt Enis

In an interview, University of Minnesota professor Brent Hecht says user-generated content (UGC) is both a critical contributor to people's Internet experience and a key resource for helping researchers extract meaning from big data. Hecht says a significant portion of search queries list Wikipedia results in the top three, in both Google and Bing. He points to a 2010 project that examined the locality of UGC and found there was disagreement in the literature. Hecht says the project determined that, whereas in the past people used to go to a city's Web page to obtain information about the city, now they usually visit the city's Wikipedia entry. "A community of people with no credentials--classic user generated content in context--is defining the way that people understand the spaces around them," Hecht observes. He also says the facts in Wikipedia articles are for the most part accurate, as Wikipedia's ability to cohere large groups of people increases the likelihood that inaccuracies will be found and corrected. However, Hecht says areas of coverage are lacking, noting the misperception that the English Wikipedia is a superset of all the other language versions. He estimates an English Wikipedia entry omits about 28 percent of the content a person would get if they could read all of the language editions.

Abstract News © Copyright 2014 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