Welcome to the May 1, 2015 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
What Is Being Learned From MOOCs? New Report Takes Stock
The Chronicle of Higher Education (04/30/15) Casey Fabris
A new report released this week lays out the current state of research into massive open online courses (MOOCs). The report is the work of the MOOC Research Initiative, which has funded 28 research projects with support from the Bill & Melinda Gates Foundation. George Siemens, an academic-technology expert at the University of Texas at Arlington and an author of the report, says the early years of MOOCs were dominated by frenzied enthusiasm for the technology that, in retrospect, seems foolish. "It's almost like we went through this sort of shameful period where we forgot that we were researchers and we forgot that we were scientists and instead we were just making decisions and proclamations that weren't at all scientific," Siemens says. The goal of the new report is to offer insight into what is known about MOOCs and how they perform in the real world. The report identifies five key research themes: student engagement and learning success; MOOC design and curriculum; self-regulated learning and social learning; social-network analysis and networked learning, and motivation, attitude, and success criteria. Siemens says student engagement is of particular interest because "a distraction is literally just a click away."
Online Voting a Step Closer Thanks to Breakthrough in Security Technology
University of Birmingham (05/01/15)
Researchers at the University of Birmingham say they have developed a new technique that could enable people to cast an election vote online, even if their home computer is infected with malware. The team, led by professor Mark Ryan, took its inspiration from security devices issued by some banks to develop a system that employs an independent hardware device used in conjunction with a personal computer. The researchers call their system Du-Vote and believe it could be ready for use in the U.K.'s 2020 or 2025 general elections. "This system works by employing a credit card-sized device similar to those used in online banking," Ryan says. "From the voter's perspective, it's straightforward: you receive a code on the device and type it back into the computer." Birmingham researcher Gurchetan Grewal says the technique is able to address the issue of a voter's computer potentially being infected by malware. The researchers say their technique can ensure security even if the security devices themselves were manufactured under the influence of a malicious actor. "This is currently the only piece of work that addresses a core problem of e-voting--namely, that someone may have viruses or other malware on their computer," Grewal says.
Supercomputing Tidal Models with the STORM Project at LSU
Inside HPC (04/28/15)
Louisiana State University (LSU) recently received a U.S. National Science Foundation grant for the STORM project, which aims to update the coastal modeling system known as Advanced Circulation (ADCIRC). ADCIRC is a multi-scale, multi-physics coastal circulation model used to recreate the effects of winds, tides, waves, and currents on large bodies of water. "The STORM project seeks to create a sustainable software framework and infrastructure that is the foundation for coastal circulation and storm surge modeling needs of a wide community for at least the next 20 years," says LSU researcher Zach Byerly. If the project is successful, ADCIRC will become a dynamic computational platform comprised of multiple solution algorithms that readily admits additional new solution algorithms and is built on a transformational new parallelization scheme that will allow scaling to at least 256,000 computer cores on modern, high-powered computing systems. The STORM project will use a next-generation parallel computing runtime system known as HPX to speed up ADCIRC, enabling better and more accurate information to be gathered more quickly and on a much larger scale.
Smartphone Secrets May Be Better Than a Password
Technology Review (04/28/15) Rachel Metz
Researchers at the Indian Institute of Technology Kharagpur, the University of Texas at Austin, and the University of Illinois Urbana-Champaign (UIUC) conducted a study designed to test how well participants could answer queries based on an activity log that included Facebook posts, websites visited, songs downloaded, and people called and texted. They collected data from participants' smartphones as well as their computers with an app, and also quizzed participants to ascertain what information they could recall. The researchers determined asking three questions concerning recent, infrequent events selected by an algorithm worked 95 percent of the time, while asking questions about other people worked less than 6 percent of the time. "Whenever there's something you and your phone share and no one else knows, that's a secret, and that can be used as a key," says UIUC professor Romit Roy Choudhury. The researchers think such a technique could supplant the expanding body of usernames and passwords most people have, or at least function as a backup for when a person cannot remember a password. They also envision the method reducing the need for password sharing for services such as Netflix. Roy Choudhury says the researchers are investigating whether their work could be useful for enterprise users.
Hottest Jobs, Industries, and Cities for IT Pay in 2015
An improving economy has led to more information technology (IT) workers reporting a raise, according to Computerworld's 2015 IT Salary Survey, which found 67 percent of respondents reported a raise and only 4 percent reported a pay cut, compared with 60 percent reporting a raise and 8 percent reporting a pay cut in 2014. In addition, total average compensation is up 3.6 percent in 2015, versus 2 percent in 2014, according to the survey, and average bonuses are up 4.6 percent. The positions of chief security officer, information security manager, and network engineer saw the greatest growth in total average compensation at 6.7, 5.3, and 5.2 percent, respectively. In terms of industry sectors, the automotive, retail trade, and wholesale trade industries saw the most growth in total average compensation at 6.8, 5.2, and 4.7 percent, respectively. Geographically, the Boston, Detroit, and San Francisco metro areas saw the largest growth in total average compensation at 5.2, 5.0, and 4.6 percent, respectively.
A Google for Handwriting
Uppsala University (04/28/15) Josefin Svensson
A digital platform launched by Uppsala University's library offers a database of digitized works from cultural heritage collections, with handwritten works rendered searchable by handwritten text recognition software. The initiative is competing with others around the world to develop a workable handwriting search application. "If someone today had an algorithm to carry out large-scale digital searches of things like the collection of manuscripts in the Vatican Library, it would be worth a fortune," notes Uppsala's Anders Brun. "Whilst the market value is enormous, so is the scale of the task." Brun and his colleagues are attempting to devise a technique that enables analysis and searches of large volumes of handwritten texts, many of which are hard to read given their age, the era, and the language in which they were written. Decoding text with a method through which the computer tries to interpret the digital image of the text forms the core of the research. "Using expert knowledge, we try to give the computer the right answer for a small portion of the material and then automate this," says Ph.D. student Frederick Wahlberg. A collaborative and interdisciplinary focus is necessary, due to the project's relevance to various humanities.
Human Brain Inspires Computer Memory
Researchers at France's Institut d'electronique fondamentale and CEA-LIST have demonstrated that physical instruments for next-generation magnetic computer memory could be capable of learning how to store and retrieve information by functioning as synapses in a neuro-inspired system. Such a system would address the problem of magnetic memory's high energy consumption. The researchers proved that probabilistic programming of magnetic tunnel junctions (MTJs) can be beneficial. They were the first to employ MTJs as connections in a system whose performance is inspired by the human brain, in that it can process very large amounts of data while expending only small amounts of energy. Probabilistic programming can enable the system to learn a function after a series of repetitions. Like the human brain's synapses, the more the MTJs are used, the greater the probability the information will be recorded. Simulations conducted by the researchers show their memory system differs from existing systems in that it can efficiently and rapidly resolve cognitive tasks such as image or video analysis while using little energy.
New UW App Can Detect Sleep Apnea Events Via Smartphone
University of Washington News and Information (04/27/15) Jennifer Langston
University of Washington (UW) researchers have developed ApneaApp, an app that uses a smartphone to wirelessly test for sleep apnea events in a user's bedroom. ApneaApp analyzes inaudible sound waves coming from the phone's speakers to track breathing patterns without needing special equipment or sensors attached to the body. During testing, the researchers say ApneaApp captured sleep apnea events with 98-percent accuracy, the same as a hospital polysomnography test. "These initial results are impressive and suggest that ApneaApp has the potential to be a simple, noninvasive way for the average person to identify sleep apnea events at home and hopefully seek treatment," says UW professor Nathaniel F. Watson. The app converts a smartphone into an active sonar system that tracks tiny changes in a person's breathing movements. The phone's speakers send out inaudible sound waves that bounce off a sleeping person's body and are detected by the phone's microphone. During nearly 300 hours of tests, ApneaApp tracked various respiratory events, including central apnea, obstructive apnea, and hypopnea with 95- to 99-percent accuracy. The tests show ApneaApp works efficiently at a distance of up to three feet, in any sleeping position, including when the user is under a blanket.
Research Advances Security and Trust in Reconfigurable Devices
Georgia Tech News Center (04/28/15) Rick Robinson
Researchers at the Georgia Institute of Technology's Georgia Tech Research Institute (GTRI) are studying security issues involving field programmable gate arrays (FPGAs). Although FPGAs are used in many applications because of the way they combine hardware performance and software flexibility, the configurability of an FPGA can be used to compromise its security. "Because FPGAs are programmable and they tightly couple software and hardware interfaces, there's concern they may introduce a whole new class of vulnerabilities compared to other microelectronic devices," says GTRI's Lee W. Lerner. One potential weakness involves side-channels, or physical properties of circuit operation that can be monitored externally. A knowledgeable attacker could analyze a device's side-channels and potentially gain enough information about its internal operations to crack encryption methods used to protect the design. However, the researchers are developing techniques to provide assurance in programmable-logic designs. For example, they are developing visualization methods that enable identifying patterns in massive logic designs that could include hundreds of thousands of nodes and connections. The researchers also developed the Trustworthy Autonomic Interface Guardian Architecture (TAIGA), a digital measure mapped onto configurable chips and wrapped around the interfaces of process controllers. TAIGA aims to establish a root-of-trust in the system, which refers to a set of functions that can always be trusted.
CMU Researchers Discover Complex Interaction Patterns on Twitter and Yelp Using Deep Learning
Carnegie Mellon News (PA) (04/29/15)
Carnegie Mellon University (CMU) researchers have developed a new way to identify the intrinsic dynamics of interaction patterns at multiple time scales. Their method involves building a deep-learning model that consists of several levels, each of which captures the relevant patterns of a specific temporal scale. The researchers say the new technique also can be used to explain the possible ways in which short-term patterns relate to long-term patterns. For example, it becomes possible to describe how a long-term pattern in Twitter can be sustained and augmented by a sequence of short-term patterns, including properties such as popularity, stickiness, contagiousness, and interactivity. The researchers used the method to build a forecasting engine they say is much more accurate than conventional approaches. Another application focuses on identifying cases in which user interactions over social media are abnormal, a process known as anomaly detection. The researchers currently are developing new methods that can use the discovered patterns to engineer the best recommended action that fits the intrinsic patterns extracted from social media. "Indeed, we want to be able to tell people how to use social media in the best possible way according to existing historical data," say CMU researchers Huan-Kai Peng and Radu Marculescu.
Picture This: Graphene Brings 3D Images Clearer and Closer
Griffith News (04/27/15) Michael Jacobson
Researchers at the Swinburne University of Technology and Griffith University have utilized the properties of graphene to create three-dimensional (3D) holographic images that could be applied to fields such as optical data storage, information processing, and imaging. "Graphene offers unprecedented prospects for developing flat displaying systems based on the intensity imitation within screens," says Qin Li from the Queensland Micro- and Nanotechnology Center of Griffith's School of Engineering. The researchers have shown that patterns of photo-reduced graphene oxide (rGO) directly written by a laser beam can produce wide-angle and full-color 3D images. "This was achieved through the discovery that a single femtosecond (fs) laser pulse can reduce graphene oxide to rGO with a sub-wavelength-scale feature size and significantly differed refractive index," according to the researchers. In addition, the spectrally flat optical index modulation in rGOs enables wavelength-multiplex holograms for full-color images. The sub-wavelength aspect is especially important because it allows for static holographic 3D images with a wide viewing angle of up to 52 degrees, according to the researchers. They say their research could revolutionize capabilities across a range of optical and electronics devices, formats, and industry sectors.
Detecting Human Life With Remote Technology
Flinders University (04/27/15)
Flinders University engineering students have developed new technology for detecting human life that they believe is faster and more efficient than published techniques. They say the technology could be used in disaster and war zones for rescue operations. As part of their master's thesis project, Laith Al-Shimaysawee and Ali Al-Dabbagh developed an algorithm that utilizes both thermal and color cameras at various resolutions. The algorithm trains their system on samples, such as images of body shapes and poses the system uses to compare against and detect real-life casualties, that are independent of the camera in use, making it more versatile. The number of training samples used is small, which makes the system very fast and results in less comparing and processing. Flinders' Nasser Asgari says the technology can be fitted to rescue robots to search for people trapped in collapsed buildings or unstable structures. In addition, he says the technology should help soldiers identify wounded people in a battlefield. Moreover, Asgari says the system is versatile enough to be used to develop a helmet or handheld device to assist emergency workers.
When Robot Personalities Mimic the Dead
Discover (04/26/15) Jeremy Hsu
The possibility of robots with personalities that mimic those of celebrities and other deceased persons could be controversial in terms of how people may react to them. Indiana University roboticist Karl MacDorman notes there may be technology already in existence to enable a robot personality to partly simulate that of a real-life person. He also says an actual person's interactions with others could provide behavioral data to inform the development of such a machine, and existing software can use vocal samples to produce a synthesized version of a person's voice. "While an individual may find comfort in having a robot or digital double impersonate a deceased loved one, others may well find this creepy, and the practice could be stigmatized," MacDorman cautions. Google was recently awarded a patent for creating robot personalities that could be downloaded as software and transferred between different robots via an online service. The patent describes that personality, or character, as "a program already stored, or it could be something in the cloud. If the latter, the robot would interact with the cloud to pull sufficient information regarding the 'new' persona to thereby recreate a simulacrum for the robot." The patent also says the robot persona's constituent information could come from a user device.
Abstract News © Copyright 2015 INFORMATION, INC.
To submit feedback about ACM TechNews, contact: [email protected]
Current ACM Members: Unsubscribe/Change your email subscription by logging in at myACM.