Association for Computing Machinery
Welcome to the May 20, 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).


Tech Giants Don't Want Obama to Give Police Access to Encrypted Phone Data
The Washington Post (05/19/15) Ellen Nakashima

A group of more than 140 technology companies, prominent technologists, and civil society groups have signed a letter addressed to President Barack Obama urging him to reject any government proposals that might allow law enforcement to force technology companies to install backdoors or otherwise weaken the encryption they use to secure their devices and data. In recent months, both the U.S. Federal Bureau of Investigation (FBI) and the Department of Justice have called for such backdoors as an increasing number of technology firms discuss making the use of strong encryption the default. Responding to plans by Google and Apple to offer strong encryption on their smartphones, FBI Director James Comey recently said he could not understand why they would "market something expressly to allow people to place themselves beyond the law." However, the signers of the letter argue there is no way to offer law enforcement a backdoor without weakening encryption in a way that could be exploited by criminals or other malicious actors. The signatories include the U.S. Public Policy Council of ACM, as well as policy experts such as SRI International Computer Science Lab principal scientist Peter G. Neumann, moderator of the ACM Risks Forum. "Strong encryption is the cornerstone of the modern information economy's security," the letter says. "This protection would be undermined by the mandatory insertion of any new vulnerabilities into encrypted devices and services."
View Full Article - May Require Free Registration | Return to Headlines | Share Facebook  LinkedIn  Twitter 

New Computer Bug Exposes Broad Security Flaws
The Wall Street Journal (05/19/15) Jennifer Valentino-DeVries

An international team of computer scientists and security researchers has discovered LogJam, an Internet bug that enables an attacker to trick a Web browser into believing it is using a regular key rather than the export version. LogJam is related to another flaw the team found over the winter, called Freak, which enabled an attacker to force another computer to use a smaller "export" key. The newly discovered weakness also could enable an attacker to read or change communications that claim to be secure. The researchers think the U.S. National Security Agency may have exploited LogJam, Freak, or other similar flaws to spy on virtual private networks. About 8 percent of the top 1 million websites are vulnerable to the new bug because they support export keys based on the same large numbers. Although Web browser makers could fix the problem by changing their programs to reject small keys, that would disable thousands of legitimate Web servers, according to the researchers. After the Freak bug was disclosed in March, browser makers agreed to reject small keys but debated where to set a threshold. In the end, the browser makers decided to move toward rejecting keys with fewer than 1,024 bits, or 309 digits, a move that could leave about 0.2 percent of secure websites inaccessible to users.
View Full Article - May Require Paid Subscription | Return to Headlines | Share Facebook  LinkedIn  Twitter 

To Handle Big Data, Shrink It
MIT News (05/20/15) Larry Hardesty

Massachusetts Institute of Technology (MIT) researchers Richard Peng and Michael Cohen have developed an algorithm that finds the smallest possible approximation of a matrix that guarantees reliable computations. In addition, for all classes of problems, the algorithm finds the approximation as quickly as possible. The researchers say the algorithm is a major improvement over previous techniques, especially for engineering and machine-learning problems. They note the algorithm is optimal for condensing matrices under any "norm." A norm is a measure of the distance between a given row of a condensed matrix and a row of the original matrix. The Manhattan distance, or 1-norm, is the first root of the sum of differences raised to the first power, and the Euclidean distance, or 2-norm, is the square root of the sum of differences raised to the second power. The new algorithm condenses matrices under the 1-norm as well as it does under the 2-norm. In addition, the researchers mathematically proved the 2-norm algorithm will yield reliable results under the 1-norm. "It's highly elegant mathematics, and it gives a significant advance over previous results," says University of California, Berkeley professor Richard Karp. Peng and Cohen will present their algorithm next month at the 47th ACM Symposium on Theory of Computing (STOC 15) in Portland, OR.

How Social Media Is Helping Nepal Rebuild After Two Big Earthquakes (05/19/15) Danielle Preiss

Several groups both in and out of Nepal have turned to crisis-mapping technology--a method of using crowdsourced information to create maps showing where aid is needed in the wake of a disaster--following a pair of massive earthquakes that have rocked the small Himalayan nation in recent weeks. Before the earthquakes, Kathmandu Living Labs, a non-profit tech company based in the Nepali capital of Kathmandu, used open data and tools such as OpenStreetMap to track development issues in the country. Following the quakes, its team began collecting Facebook posts and tweets to generate a single organized quake map that it shared with several relief groups and the Nepal Army. Another group called Sankalpa also has been creating maps, which it based on information gathered from mobile messaging service Sparrow SMS. Sankalpa's maps help to show where and what kinds of aid are being requested. They also help to cut down on inaccurate reporting. However, due to the tenuous state of the phone and data networks in Nepal, such efforts largely have been restricted to more urban areas. In addition, even when maps help to pinpoint where aid is needed, getting that aid to the people there often remains the hardest part of the process.

American Innovation Lies on Weak Foundation
The New York Times (05/19/15) Eduardo Porter

The basic scientific research that makes possible the great innovations of U.S. companies is being threatened by falling funding from both the public and private sectors. According to the U.S. National Science Foundation, investment in research and development (R&D) as a share of the U.S. economy has flatlined in recent years, stabilizing at about 2.9 percent of gross domestic product. Although this is not far from the overall peak, other countries are pulling ahead of the U.S. even as investment in basic research has steadily shrunk over the last decade. Of particular concern is falling government investment in basic research, which has become a target for cost-cutting in recent years. Meanwhile, the loss of federal funding for basic research is not being compensated for by corporate research. The era of Bell Labs and Xerox PARC, which were once responsible for innovations such as the transistor and the graphical user interface, is largely over. Although businesses are pursuing patents more aggressively than ever, they appear less willing to invest in the basic research that underlies those patents, in part because of an increased focus on short-term gains in the corporate world. The drop in R&D threatens continuing productivity growth, damages the career prospects of young scientists, and discourages young Americans from pursuing scientific careers.
View Full Article - May Require Free Registration | Return to Headlines | Share Facebook  LinkedIn  Twitter 

Computing at the Speed of Light
University of Utah News (05/18/15) Vincent Horiuchi

University of Utah researchers have developed an ultracompact beamsplitter for dividing light waves into separate channels for information, a breakthrough say they could lead to the next generation of computers and mobile devices that can operate at speeds millions of times faster than conventional machines. The researchers say the device is a step toward the development of silicon photonic chips that compute and shuttle data with light rather than electrons. Although light is the fastest medium that can be used to transmit information, that information must be converted to electrons before a device can manipulate it. However, that bottleneck could be eliminated if the data stream remained as light within computer processors. The Utah researchers achieved that process by creating a much smaller form of a polarization beamsplitter on top of a silicon chip that can split guided incoming light into its two components. On a future silicon chip, the beamsplitter would be just one of several passive devices used to direct light waves in different ways. By shrinking the devices down to just 2.4-square microns, researchers will be able to squeeze millions of them on a single chip. The researchers note their design would be inexpensive to produce because it uses existing fabrication techniques for creating silicon chips.

New Chip Architecture May Provide Foundation for Quantum Computer
Georgia Tech News Center (05/18/15) John Toon

Georgia Institute of Technology (Georgia Tech) researchers have demonstrated a device that enables more electrodes to be placed on a chip, a breakthrough they say could help increase qubit densities and bring the field one step closer to a quantum computer that can perform complex algorithms. Qubits, which exploit a quantum property called superposition, can be correlated with each other in a way that classical bits cannot, allowing for a new kind of massively parallel computation, but only if many qubits at a time can be produced and controlled. The challenge is scaling the technology up into a useful device. One promising qubit candidate is individual ions trapped inside a vacuum chamber and manipulated with lasers. However, the scalability of modern trap architectures is limited because the connections for the electrodes needed to generate the trapping fields come at the edge of the chip, which means their number is limited by the chip perimeter. The Georgia Tech approach uses new microfabrication techniques that allow more electrodes to fit onto the chip while preserving the laser access required. The researchers also freed up more chip space by replacing area-intensive surface or edge capacitors with trench capacitors and strategically moving wire connections.

Key Strategies Can Boost Donations at Crowdfunding Sites, Stanford Experts Say
Stanford Report (05/18/15) Tom Abate

Stanford University researchers have shown how crowdfunding websites can use data science to boost the cash value of donations, confirming along with other findings the importance of a timely "thank you." The researchers analyzed millions of online donations collected over a 14-year period. The findings suggest increasing the number of repeat givers by just 10 percent could yield more than 60 percent more money donated over time. "The people who become repeat givers tend to give more with successive donations, and they also recruit others to give," says Stanford graduate student Tim Althoff. The research also found ways to increase donor retention, such as posting smaller requests, as well as how to time "thank yous" and project reports. "In this case we had excellent data that enabled us to discover new tools and techniques for the fundraising community," says Stanford professor Jure Leskovec. The researchers estimated the potential to increase fundraising based on boosting donor retention using a model to project what would be the result of pushing donor retention from 26 to 36 percent. The projection showed a 60-percent increase in total amount donated was likely as repeat givers tend to increase their gifts and also recruit their friends.

A Nano-Transistor Assesses Your Health Via Sweat
Swiss Federal Institute of Technology in Lausanne (05/15/15) Laure-Anne Pessina

Swiss Federal Institute of Technology in Lausanne (EPFL) researchers have developed an ultra-low-power sensor that enables real-time scanning of the contents of liquids such as perspiration. The researchers say the sensor is made using state-of-the-art silicon transistors and is compatible with advanced electronics and is accurate enough to be used with mobile sensors that monitor health. "Our technology detects the presence of elementary charged particles in ultra-small concentrations such as ions and protons, which reflects not only the pH balance of sweat but also more complex hydration of fatigues states," says EPFL researcher Adrian Ionescu. The researchers fixed a microfluidic channel, through which the fluid to be analyzed flows, to the transistor. The technology relies on a layered design that isolates the electronic part from the liquid substance. "In our chip, sensors and circuits are in the same device--making it a 'sensing integrated circuit,'" says EPFL researcher Sara Rigante. "This proximity ensures that the signal is not disturbed or altered. We can thereby obtain extremely stable and accurate measurements." The researchers note the transistor's size of just 20 nanometers makes it possible to place a whole network of sensors on one chip, with each sensor locating a different particle.

New Mobile App Expands the Outreach of SAWBO Videos
University of Illinois News Bureau (05/15/15) L. Brian Stauffer

The University of Illinois' SAWBO initiative, which develops animated educational videos, recently launched the Deployer, a free app that enables users to view, download, and share SAWBO's video library over Bluetooth connections. SAWBO's collection of two- and three-dimensional animations currently covers about 50 topics, and the Deployer app enables users to scroll through and filter the list by topic, language, or country. "Individuals around the world can download our videos through the app when they have Internet access, store the animations on their phones, and share them with other devices whenever and wherever they want," says Illinois professor Barry Pittendrigh. The initiative's goal is for Deployer to be used as a one-step, automated system of transferring knowledge from experts directly into the field. The Deployer software also enables SAWBO researchers to track the geographic locations where the app and videos are downloaded or shared. The app currently is only available in English, but Pittendrigh says the researchers are developing versions in Amharic, French, Portuguese, and Spanish.

The Machine Vision Algorithm Beating Art Historians at Their Own Game
Technology Review (05/11/15)

Recent advances in machine learning such as deep convolutional neural networks are enabling the development of machines that perform pattern-recognition tasks. Rutgers University researchers Babak Saleh and Ahmed Elgammal used machine-learning techniques to train algorithms to recognize the artist and style of fine-art paintings. The researchers began with a database of images of more than 80,000 paintings by more than a 1,000 artists spanning 15 centuries and covering 27 styles. They took a subset of the images and used them to train machine-learning algorithms to identify certain features, such as overall color and describing the objects in the image. The end result is a vector-like description of each painting that contains 400 different dimensions. The researchers tested the algorithms on a set of unfamiliar paintings and found they achieved 60-percent accuracy in identifying the artist in the paintings it saw and 45-percent accuracy in identifying the style. The machine-learning approach also can pick up similarities, such as linking expressionism and fauvism, within just a few months. A potential application of the new algorithms is identifying paintings with similar characteristics to detect potential influences between artists.

PolyU Develops Novel Computer Intelligence System for Acute Stroke Detection
Hong Kong Polytechnic University (05/11/15)

Hong Kong Polytechnic University scientists have developed a computer-aided detection system for acute stroke using computer-intelligence technology. Within just three minutes, the system analyzes 80 to 100 computer images to detect if the patient underwent ischemic stroke or hemorrhagic stroke, with a detection accuracy of 90 percent. The researchers say the system could help minimize damage through timely diagnosis and treatment within three hours of stroke onset. The first part of the computer-aided detection (CAD) stroke technology is an algorithm for automatic extraction of areas of suspected region of interest, and the second part is an artificial neural network that classifies regions of interest. The CAD stroke computer "learns" the defining features of stroke, and performs automated reasoning. Computerized tomography scans are fed into the CAD stroke computer, which makes calculations and comparisons to locate areas suspected of insufficient blood flow. It detects where the images look "abnormal" due to factors such as loss of insular ribbon or sulcus and dense middle cerebral artery, which are highlighted for doctors' review.

MenuetOS, an Operating System Written Entirely in Assembly, Hits 1.0
Computerworld Australia (05/15/15) Rohan Pearce

After almost 15 years of development, MenuetOS, an x86-based operating system (OS) written entirely in assembly language, has reached version 1.0. The OS comes with a long list of features, but is still compact enough that it could, conceivably, fit on a floppy disk. Menuet's features include support for preemptive multitasking and multiple processors, as well as for a variety of networking, audio, and USB-based devices, including mice and keyboards, TV tuners, webcams, and printers. It also sports a robust graphical user interface that includes support for transparency. The OS comes with range of applications, including a media player, Web browser and server, an FTP client and server, a chess program and Tetris clone, and an email client. "Menuet's primary goal was to write an operating system 100 percent in assembly for faster code execution and smaller memory use," says Ville Turjanmaa, Menuet's creator and project leader, who notes another goal was to include common features "we would expect from a modern operating system." Turjanmaa says the OS is well-suited to tasks such as precise timing and machine control. He says future versions of the OS will focus on improving various application classes, as well as "other interesting plans as well, but it's too early to talk about those."

Abstract News © Copyright 2015 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