Welcome to the February 14, 2020 edition of ACM TechNews, providing timely information for IT professionals three times a week.
Please note: In observance of the U.S. Presidents' Day holiday, TechNews will not be published on Monday, Feb. 17. Publication will resume Wednesday, Feb. 19. ACM TechNews mobile apps are available for Android phones and tablets (click here) and for iPhones (click here) and iPads (click here).
To view "Headlines At A Glance," hit the link labeled "Click here to view this online" found at the top of the page in the html version.
The online version now has a button at the top labeled "Show Headlines."
|
|
|
MIT Researchers Say Mobile Voting App is Rife with Vulnerabilities
Computerworld
Lucas Mearian
February 13, 2020
Massachusetts Institute of Technology researchers found the mobile voting Voatz app piloted in the U.S. contains bugs that hackers could exploit to alter, block, or expose how users are voting. The Voatz platform combines biometrics like cellphone-based facial recognition and hardware-backed keystores to deliver end-to-end encrypted and voter-verifiable ballots, and employs blockchain as a tamperproof electronic ledger to store voting results. Critics of mobile or online voting warn the app opens up the possibility of attacks associated with infiltrating voters' systems with malware, or infecting computers in the elections office that manage and count ballots. Jeremy Epstein of ACM's U.S. Technology Policy Committee said, "Any election official using Voatz products would be well advised to cancel their plans, before a stealthy attack in a real election compromises democracy."
|
Researchers Devise Approach to Reduce Biases in Computer Vision Datasets
Princeton University
Molly Sharlach
February 12, 2020
Computer scientists at Princeton and Stanford universities have developed ways to produce less-biased datasets containing images of people, and propose applying these methods to the ImageNet computer-vision training database. The researchers systematically identified non-visual concepts and offensive categories like racial and sexual characterizations in ImageNet's person categories, and proposed their deletion. They also devised a tool that lets users specify and retrieve image sets of people balanced by age, gender expression, or skin color to enable algorithms to classify faces and activities in images more fairly. Said Princeton’s Olga Russakovsky, “Computer vision now works really well, which means it’s being deployed all over the place in all kinds of contexts. This means that now is the time for talking about what kind of impact it’s having on the world and thinking about these kinds of fairness issues.”
|
Data Driving New Approaches to Transportation
The New York Times
Norman Mayersohn
February 10, 2020
Shifts in transportation highlight cities' need for data management to cope with traffic, digital platforms, and other trends. For example, Los Angeles uses data sent from dockless electric scooters to scooter fleet owners, who have to know the vehicles' locations in order to collect them each night for charging, and reposition them for the next morning when demand peaks. Knowing historical scooter routes also helps policymakers plan infrastructure that addresses congestion. The open source Mobility Data Specification software platform compiles this data stream in a universally readable format. Meanwhile, startups like Lacuna Technologies design software to support open source systems for transportation agencies, with the goal of accommodating future trends like three-dimensional traffic systems that include delivery drones and air taxis.
*May Require Paid Registration
|
What's the Hottest Job in Tech?
IEEE Spectrum
Tekla S. Perry
February 7, 2020
Technology recruiting firm Dice's analysis of 6 million U.S. job postings last year found the single largest category of tech job openings was that of software developer, while data engineering jobs saw the fastest growth in job postings. Dice found 12% of all tech job listings in 2019 were for software developers, followed by network engineers and systems engineers. Demand for data engineers rose 50% last year over 2018 levels, with Structured Query Language and the open source Kubernetes platform respectively representing the most in-demand and fastest-growing skills requested in job postings. Nearly 20% of developer job postings specified familiarity with the Python programming language as a desired skill, as did 75% of data scientist listings and 64% of data engineer listings.
|
Making the Internet More Energy Efficient
Chalmers University of Technology
February 13, 2020
Researchers at Sweden’s Chalmers University of Technology designed optimized error-correcting data chip circuits to reduce their power consumption and boost the Internet's energy efficiency. The chips consume about a tenth the energy that conventional processors use. The researchers also demonstrated the benefits of using optical frequency combs in fiber-optic communication systems rather than having separate laser transmitters for each frequency channel; the combs simultaneously transmit light at all wavelengths, making signal reception more energy efficient. Said Chalmers' Erik Agrell, "Improving the energy efficiency of data transmission requires multidisciplinary competence. The challenges lie at the meeting points between optical hardware, communications science, electronic engineering, and more."
|
Simulations Identify Missing Link to Determine Carbon in Deep Earth Reservoirs
University of Chicago
Emily Ayshford
February 10, 2020
Collaborators at the University of Chicago (UChicago) Pritzker School of Molecular Engineering and Hong Kong's University of Science and Technology have created a computer model to help measure carbon levels in the deep reservoirs of the Earth's mantle. While more than 90% of the Earth’s carbon, according to some estimates, is buried in its interior, it can influence the concentration of carbon near the surface, which can affect climate change. UChicago's Giulia Galli and Ding Pan developed a technique that combines spectroscopy results with calculations based on quantum mechanics to quantify carbon concentrations in water. Said Galli, “Our computational strategy will greatly facilitate the determination of the amount of carbon at the extreme conditions of the Earth’s mantle."
|
Researchers Entangle Quantum Memory at Facilities Over 50 km Apart
Ars Technica
John Timmer
February 12, 2020
Researchers in China at the University of Science and Technology of China and other institutions have entangled quantum memory at facilities separated by more than 20 kilometers (12 miles) and 50 kilometers (31 miles), although the memory suffered decoherence in the second instance. The process uses clouds of about 100 million cold atoms that can be treated as a single quantum object, with somewhat greater stability from distributing the quantum state evenly across all the atoms. While the researchers acknowledged losses from filtering noise and sending photons into the fiber, they said the entire process is more than 30% efficient, end to end.
|
Printing Tiny, High-Precision Objects in Seconds
EPFL (Switzerland)
Sarah Perrin
February 13, 2020
Researchers at the Swiss Federal Institute of Technology in Lausanne (EPFL) have developed a high-precision technique for three-dimensionally (3D) printing small, soft objects in seconds. The method employs the principles of tomography, in which models of objects are constructed from surface scans. The printer transmits a laser through a translucent gel that is either organic or liquid plastic, hardening the material as algorithms calculate the areas the laser targets, the beam's angles, and intensity. The system currently produces 2cm structures with 80-micrometer precision, and new devices should be able to print larger objects, potentially up to 15 centimeters. The researchers partnered with a surgeon to test 3D-printed arteries fabricated with this method, and the technology could potentially have bioprinting applications due to its ability to print solid objects of different textures.
|
Mac Software Threats Climbed 400% in 2019, More Than Microsoft Windows, Report Says
USA Today
Edward C. Baig
February 12, 2020
A report from cybersecurity company Malwarebytes concluded that Mac computers are more vulnerable to software threats than Microsoft Windows systems. The report observed greater growth in Mac-targeting adware and potentially unwanted programs compared to Windows, and threats to Mac PCs increased 400% year-over-year in 2019. Meanwhile, Mac detections per endpoint increased from 4.8 in 2018 to 11 in 2019, which was nearly twice as many for Windows. Mac threats also topped Malwarebytes' overall threat detections list for the first time, with the NewTab and PCVARK exploits ranking second and third, respectively, on the list of the most prevalent threat detections across all platforms. Malwarebytes suggested Macs may be more appealing targets due to their growing market share, while iPhones may be susceptible to iOS malware because there is no way "to scan for it."
|
Computer-Based Weather Forecast: Algorithm Outperforms Mainframes
Johannes Gutenberg University of Mainz (Germany)
February 13, 2020
Researchers at Johannes Gutenberg University Mainz in Germany and Switzerland's University of Lugano (USI) have created an algorithm that solves complex problems on personal computers, while consuming less processing power than a supercomputer. The researchers based the Scalable Probabilistic Approximation (SPA) algorithm on the Lego principle, by which complex systems are deconstructed into discrete states or patterns, making only a few patterns necessary to analyze large datasets and predict their future behavior. The researchers used the algorithm to forecast surface temperatures in Europe one day ahead, found its error rate was 40% lower than that of systems typically used by weather services. Said USI's Illia Horenko, "The process is easier and cheaper and the results are also better compared to those produced by the current state-of-the-art supercomputers."
|
Unifying Human Conflict Using Data Science
GW Today
Kristen Mitchell
February 11, 2020
A George Washington University (GW) researcher used computational modeling to discover a hidden unified mathematical pattern underlying human conflicts throughout history. GW's Neil Johnson and colleagues followed a process established by scholar Lewis Fry Richardson for measuring the distribution of conflict sizes based on total causalities, to assess individual clashes within conflicts and terrorist attacks and uncover so-called power laws. Watching his son play multiplayer video games inspired Johnson to consider a connection between two power-law patterns, suggested by apparent shifts in the numbers for real-time casualties for clashes during missions, and the aggregate score for entire missions. Computational simulation confirmed this relationship. Johnson thinks data scientists should rethink their approach to structuring data so such an overall picture will be more evident.
|
In the Eternal Quest to Decode Fedspeak, Here Come the Computers
The Wall Street Journal
James Hookway
February 9, 2020
Investors are using artificial intelligence to decrypt the purposefully vague or baffling jargon used by central banks to predict their financial policies. The Royal Bank of Canada's Peter Schaffrik co-developed the ECB-O-Meter algorithm, which scans speeches by European Central Bank leaders to predict potential future strategies. Meanwhile, U.S.-based Prattle Analytics tracks the U.S. Federal Reserve and other central banks by aggregating keywords for each bank, then applying machine learning to monitor asset prices when officials use those words in public comments. University of Tokyo programmers algorithmically analyzed facial expressions at televised news conferences for hints of future policy, noting Bank of Japan Governor Haruhiko Kuroda's looks of anger and disgust foreshadowed the bank's rollout of negative interest rates in 2016. Some developers expect machine models to overtake humans, although Schaffrik said humans must occasionally step in to feed algorithms new data.
*May Require Paid Registration
|
AI Shortcuts Speed Simulations Billions of Times
Science
Matthew Hutson
February 12, 2020
Researchers led by the University of Oxford in the U.K. used artificial intelligence to generate accurate machine learning emulator algorithms for accelerating simulations billions of times, for all scientific disciplines. The neural network-based emulators absorb the inputs and outputs of a full simulation, seeking patterns and learning to guess what the model would do with new inputs while avoiding the need to run the full simulation many times. The Deep Emulator Network Search (DENSE) method randomly inserts computation layers between network inputs and outputs and trains the system with the limited data, so added layers that improve performance are more likely to end up in future variations. DENSE-produced emulators for 10 simulations in physics, astronomy, geology, and climate science were 100,000 to 2 billion times faster than the models with the addition of specialized graphical processing chips—and were highly accurate.
|
|
