Association for Computing Machinery
Welcome to the October 8, 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).

HEADLINES AT A GLANCE


Tech Companies Hope to Introduce Coding to 100 Million Students
The Wall Street Journal (10/08/14) Jeff Elder

The CEOs of two dozen major tech companies, including Google and Microsoft, will announce their support on Wednesday for a project by nonprofit Code.org that seeks to introduce computer science to 100 million students worldwide. The companies will promote Code.org's Hour of Code campaign, which encourages students to explore computer coding through hour-long online tutorials. The support will take the form of encouraging their employees to try out Hour of Code tutorials and encourage students to do the same during Computer Science Education Week this December. The companies also will encourage their employees to contribute to an Indiegogo crowdfunding campaign that seeks to raise $5 million to help train teachers in providing computer science education. The effort is an attempt to help encourage broader participation in computer science among women and minority communities as well as to meet anticipated growing demand for workers with computer science skills. ACM and CSTA are major partners in Code.org. Other companies supporting Hour of Code include Disney, Dropbox, Eventbrite, GoDaddy, Salesforce.com, Target, Yelp, and Zillow. "Some of these companies are competitors," says Code.org CEO Hadi Partovi. "We represent a nice, safe place for them to channel their efforts."
View Full Article - May Require Paid Subscription | Return to Headlines | Share Facebook  LinkedIn  Twitter 


One in Three Jobs Will Be Taken by Software or Robots by 2025
Computerworld (10/06/14) Patrick Thibodeau

By 2025, software, robots, or other “smart machines” will have replaced as many as one in three existing jobs, Gartner predicts. Gartner research director Peter Sondergaard says smart machines represent an emerging "super class" of technologies that can perform a wide variety of work, from physical tasks to intellectual work, and will include drones and robots performing work such as pipeline maintenance and crop dusting to computers and software grading essays and multiple choice tests. "Knowledge work will be automated," Sondergaard says. "New digital businesses require less labor; machines will be make sense of data faster than humans can." He predicts financial analysis, medical diagnostics, and a wide variety of data analysis jobs will be automated away in the near future. Nuverra Environmental Solutions CIO Lawrence Strohmaier says these jobs will not be destroyed so much as they will be replaced by different jobs. "The shift is from doing to implementing, so the doers go away but someone still has to implement," Strohmaier says. Both he and Gartner analyst David Aron say the rise of smart machines will be accompanied by an increase in the number of IT workers to operate and maintain them.


The Ethics of Hacking 101
The Washington Post (10/07/14) Ellen Nakashima

Some of the U.S.'s most prestigious cybersecurity university programs make a point of teaching their students offensive skills, but in doing so also must address the thorny ethical issues surrounding the gray-hat world. At the University of Tulsa, students who mostly go on to careers with government agencies such as the U.S. Central Intelligence Agency and the National Security Agency are taught to hack into oil pipelines and energy infrastructure and carry out penetration testing for clients, while Carnegie Mellon University has developed an elite team of hackers who dominated this year's capture-the-flag hacking competition at Defcon in Las Vegas. Professors from both universities say they take great pains to teach ethics to their students and ensure they are not teaching offensive skills to individuals who might use them for malicious purposes. Tulsa professor Sujeet Shenoi says all of his students are subjected to rigorous background checks and interviews and go into the program with the understanding they will go on to government work. Still, it is unclear how certain any educational institution can be of the integrity of its students and to what extent it can influence their ethical outlook.
View Full Article - May Require Free Registration | Return to Headlines | Share Facebook  LinkedIn  Twitter 


A Peek Inside the Internet's Favorite File-Sharing Network
Science (10/06/14) Jia You

A new study is among the first to take a deep look at the behavior of users of the BitTorrent peer-to-peer file-sharing network. Rovira i Virgili University's Jordi Duch and colleagues collaborated with computer scientists at Northwestern University, the makers of BitTorrent plug-in Ono, which accelerates BitTorrent download speeds in exchange for allowing researchers to monitor users' download behaviors. The plug-in anonymizes users' data and does not track the content of their downloads. However, Duch and his team were able to track what sorts of media users were downloading by noting that certain file sizes tended to correlate with specific media. The researchers examined about 10,000 BitTorrent users for four years and drew several conclusions. For example, users tended to be very consistent in the types of media they shared and often downloaded on a weekly basis. They also found that users in a given country had similar download behaviors, with users in nations with more developed Internet service offerings favoring music downloads while those in less developed nations favored movies and TV shows. Duch says this is likely because the presence of services such as Netflix have given users in developed countries an inexpensive and legal means of accessing such content.


Printing in the Hobby Room: Paper-Thin and Touch-Sensitive Displays on Various Materials
Saarland University (10/07/14)

Computer scientists in Saarland University's Embodied Interaction research group say they have developed inexpensive technology that eventually could enable users to print paper-thin and touch-sensitive displays on all types of materials. For example, the researchers have demonstrated a postcard depicting an antique car, with two segments on a flexible display, and the back axle and steering wheel rim light up when a button is pressed. Printed with an off-the-shelf printer, the postcard is electro-luminescent and emits light when connected to electric voltage, similar to a car dashboard. The process involves designing a digital template with programs such as Microsoft Word or PowerPoint for the display, and using the team's technology to print the templates. The researchers say their method can be used to print and integrate displays on everything from furniture and decorative accessories to bags and wearable items. They say one possible application is a wristwatch strap that lights up when a text message is received. "If we combine our approach with 3D printing, we can print three-dimensional objects that display information and are touch-sensitive," says Saarland researcher Jurgen Steimle.


Crumpled Graphene Could Provide an Unconventional Energy Storage
MIT News (10/03/14) David L. Chandler

Massachusetts Institute of Technology (MIT) researchers have found crumpling a piece of graphene "paper" can lead to new properties that could be helpful in the creation of extremely stretchable supercapacitors to store energy for flexible electronic devices. The new, flexible supercapacitors should be easy and inexpensive to fabricate, according to the researchers. They have shown that by crumpling a sheet of graphene paper into a chaotic mass of folds, a supercapacitor can be bent, folded, or stretched to as much as 800 percent of its original size. The material also can be crumpled and flattened up to 1,000 times without significant performance degradation. "The graphene paper is pretty robust and we can achieve very large deformations over multiple cycles," says MIT professor Xuanhe Zhao. The researchers generated the graphene paper by placing a sheet of the material in a mechanical device that first compressed it in one direction, creating a series of parallel folds, and then in the other direction, resulting in a chaotic, rumpled surface. Stretching the material causes the folds to smooth themselves out. The researchers note this method also can be used for other applications, such as creating an electrode in a flexible battery or a stretchable sensor.


Twitter Data Mining Reveals America's Religious Fault Lines
Technology Review (10/06/14)

Wright State University researchers have analyzed more than 250,000 U.S. Twitter users who have declared an affiliation with various religions, providing insights into the nature of religious activity on Twitter and how these groups of self-declared users differ from each other. The researchers began by filtering the biographies of Twitter users by analyzing keywords associated with seven different religious groups, including atheists, Buddhists, Christians, Hindus, Jews, Muslims, and undeclared. The researchers found Christians to be the largest group with more than 200,000 users, while Hindus were the least represented group with 200 users. The researchers used the data to calculate the religiousness of each U.S. state and compared the results to a Gallup survey of religiousness. "They agree on 11 of the top 15 most religious states (e.g., Alabama, Mississippi, and South Carolina) and 11 of the top 15 least religious states (e.g. Vermont, New Hampshire, and Massachusetts)," the researchers note. "If our observations were to hold in a broader context, it could be seen as good for society that followers of religious groups differ most in references to religious practice and concepts, rather than in everyday aspects such as music, food, or other interests," says Wright State Ph.D. student Lu Chen.


Women Talk Tech and Science at MIT Panel
The Boston Globe (10/06/14) Erin Connolly

The Massachusetts Institute of Technology (MIT) recently hosted a panel on women in science, technology, engineering, and mathematics (STEM) fields, providing an opportunity for women to discuss their experiences in the tech industry. Many of the women expressed feeling discouraged from even pursuing technology in school and feeling excluded once they did break into the industry. "In high school, I was told [by my guidance counselor and my teachers] to pursue something that was a little bit less demanding," says Intel's Gabriela Gonzalez. The panelists also said women in STEM careers may feel pressure to present themselves in a certain way just to be acknowledged or recognized. "As a woman I really had to go in thinking, 'What do men act like, how should I act like a man but not too much like a man so they don't think I'm strange,'" says MIT Ph.D. student Jean Yang. "When you spend all your time thinking about that, it's really hard to do much else." Even some MIT students say they have felt unwelcome on campus at first. "You have to work really hard for people to take you seriously," says MIT student Tami Forrester. "We really need to make sure that women have the privilege to move forward in any field we want."


Computer Says "Try This"
The Economist (10/04/14)

More than 90 groups of researchers currently are developing hypothesis generation software that can be used on a wide range of scientific inquiries. For example, a French National Institute for Agricultural Research team is working on Methode d'Inference, software that analyzes data on hormones and the 1,500 types of receptor molecules with which they interact. Although the program is still a work in progress, the researchers say it already has prevented the duplication of work within the institute and produced a novel hypothesis about the mode of operation of follicle-stimulating hormone. Meanwhile, Baylor College of Medicine researchers are developing hypothesis generation software that searches for proteins called kinases that activate another protein, p53, which curbs the growth of cancers. The researchers used the software to read the abstracts of 186,879 papers and produced a list of the most promising kinases for experiments. In addition, University of California, San Diego researchers are developing brainSCANr, a program designed to help neuroscientists choose research projects.


Pressing the Accelerator on Quantum Robotics
Plataforma SINC (Spain) (10/02/14)

Quantum tools should revolutionize robots, automatons, and other agents that use artificial intelligence (AI), according to researchers from the Complutense University of Madrid (UCM) and the University of Innsbruck. They have confirmed that quantum machines will be able to adapt better and faster to the environment surrounding them. The theoretical work focuses on applying quantum computing to machine learning and the creation of highly accurate models and predictions. "Building a model is actually a creative act, but conventional computers are no good at it," says UCM's Miguel A. Martin-Delgado. "That is where quantum computing comes into play. The advances it brings are not only quantitative in terms of greater speed, but also qualitative: adapting better to environments where the classic agent does not survive. This means that quantum robots are more creative." Martin-Delgado and fellow UCM researcher G. Davide Paparo say the results of their research represent "a step forward towards the most ambitious objective of artificial intelligence: the creation of a robot that is intelligent and creative, and that is not designed for specific tasks." The research comes under a new discipline called quantum AI.


DARPA Delving Into the Black Art of Super Secure Software Obfuscation
Network World (10/03/14) Michael Cooney

The U.S. Defense Advanced Research Projects Agency (DARPA) wants to build cryptographic and algorithmic security techniques, ideas, and concepts into programs to make them nearly unbreakable. For a new program called SafeWare, DARPA wants to research and develop new mathematical foundations and implementation paths for provably secure program obfuscation. Software obfuscation focuses on making the important underlying code as untouchable as possible to intruders looking to access its information. "The goal of the SafeWare research effort is to drive fundamental advances in the theory of program obfuscation and to develop highly efficient and widely applicable program obfuscation methods with mathematically proven security properties," DARPA says. For example, the agency is looking for methods that have general-purpose applicability to standard, non-pathological program types and do not rely on special hardware or physical resources. Moreover, the techniques DARPA is seeking must not be substantially diminished in effectiveness even if they are fully understood by the enemy. On the other hand, DARPA is not interested in methods that require encrypted hardware, distributed computation, or distributed storage, or methods that require exotic physical states or resources not found on commodity digital hardware.


Carnegie Mellon Leads New NSF Project Mining Educational Data to Improve Learning
Carnegie Mellon News (PA) (10/01/14) Byron Spice

Carnegie Mellon University (CMU) will spearhead a five-year early implementation project to enhance educational outcomes and promote the science of learning via LearnSphere, a distributed infrastructure designed to securely store data on how students learn. LearnSphere is sponsored by the U.S. National Science Foundation (NSF) under the Data Infrastructure Building Blocks program. LearnSphere will enable course developers and instructors to access more than 550 datasets from interactive tutoring systems, educational games, and massively open online courses (MOOCs) so they can improve education and learning via data-driven course design. In addition, mining the data contained in LearnSphere will help researchers draw deeper insights about the learning process. The project will develop a graphical interface integrating data sources and analytic tools. NSF's Irene Qualters says the LearnSphere project builds on the Pittsburgh Science of Leaning Center's DataShop repository, which played a key role in the expansion of the educational data-mining discipline. LearnSphere project leader and CMU professor Ken Koedinger notes most of DataShop's content is derived from heavily interactive systems, which LearnSphere will supplement with additional data, such as information concerning student behavior and performance in MOOCs. CMU professor Carolyn Rose says LearnSphere can help improve student learning with personalized, just-in-time support.


Fueled by Algorithms
Pacific Northwest National Laboratory (10/03/14)

An international team of researchers has examined the maximum bipartite matching problem using existing combinatorial optimization algorithms designed to find the best solution with minimum enumeration. The researchers say the study involved exploiting modern multi-core computers, and provided a new parallel version of the push-relabel algorithm for bipartite graph matching that works well for shared-memory computing systems. The research also included an examination of the algorithmic performance, showing viable and improved scaling on various multi-core machines. "We are exploring a new class of algorithms for maximum matching that have direct implication for other algorithms, such as network flows," says Pacific Northwest National Laboratory researcher Mahantesh Halappanavar. As part of the research, the team tested several techniques to expedite the computation of maximum matching, including using greedy initialization algorithms, search-space pruning techniques, and switching to serial computation when the algorithm runs out of concurrency. The researchers then compared the method with a separate class of algorithms based on the standard technique of augmentation. In the future, the team wants to expand the research to include preflow-push algorithms used for computing maximum flows.


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