Welcome to the December 21, 2020 edition of ACM TechNews, providing timely
information for IT professionals three times a week.
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."
|
|
|
Programming Language Security: These Are the Worst Bugs for Each Top Language
ZDNet
Liam Tung
December 15, 2020
Veracode, a static code analysis security firm, scanned 130,000 apps for security issues and identified the main vulnerabilities for which developers need to be concerned. The firm revealed that for apps written in JavaScript and PHP, 31.5% and 74.6%, respectively, have at least one cross-site scripting flaw. Seventy-one percent of PHP apps also have cryptographic issues. Information leakage is the main issue for apps written in .NET (showing up in 62.8% of apps scanned), while error handling is the most frequent issue for apps written in C++ (66.5%). Carriage Return or Line Feed injection was found in 64.4% of Java apps, and cryptography issues were present in 35% of Python apps. High-severity flaws were found in 59% of C++ apps and 52% of PHP apps, but just 24% of Java apps and 9.6% of JavaScript apps. Veracode's Chris Eng said, "When we look at the overall numbers, as an industry we haven't eradicated any category of flaw over the past 10 years."
|
Social Work Faculty Uses AI to Fight Substance Abuse in Youth Experiencing Homelessness
University of Denver
Lorne Fultonberg
December 10, 2020
Researchers at the University of Denver, the Pennsylvania State University (Penn State), and the University of Southern California (USC) studied more than 1,600 youths in six states and used algorithms and machine learning to interpret the resulting data to determine how and when to intervene with regard to substance use disorder (SUC) in homeless youth. Penn State's Amula Yadav developed the Comprehensive Opioid Response Tool Driven by Artificial Intelligence (CORTA), which considers multiple factors to predict one???s likelihood of developing an opioid addiction. USC's Aida Rahmattalabi developed the Group-based Intervention Decision aid (GUIDE) to assess an individual's social network to develop more effective intervention and recovery groups. The researchers indicated that the use of artificial intelligence could help social workers identify youths at high risk of SUC and facilitate appropriate interventions.
|
OrCam's AI-Based Reading Device Wins CES Innovation Award
The Times of Israel
Shoshanna Solomon
December
16, 2020
Israeli artificial intelligence (AI) device developer OrCam Technologies has been named to receive the CES 2021 Innovation Awards Best of Innovation award in the ???accessibility??? category for its OrCam Read digital reader, a pen-sized assistive handheld device developed for people with reading challenges. The device is designed to read aloud full pages or screens of text from any printed or digital surface, without an Internet link. The device???s AI algorithms enable intuitive point-and-click function, triggering two precision laser-guidance options to read highlighted text or target a point to start reading. Its text-to-speech reading engine also has an interactive feature integrating computer vision and natural language understanding as a voice-activated assistant.
|
Novel Data Storage System Ready to Offer Relief for Frustrated Users, Companies
Purdue University Research Foundation News
Chris Adam
December 16, 2020
Purdue University researchers have developed a solution designed to cut download wait times and boost storage efficiency. Purdue's Vaneet Aggarwal said, "Our system provides a big advantage in that it reduces latency to provide a better experience for users and for the organization or company." The Purdue team's caching framework incorporates an algorithm that optimizes caching across a distributed storage system, which enhances performance by splitting data across multiple servers. As a result, ???chunks??? of data can be stored separately, but closer to end-users, which reduces congestion in the network, as well as delay times.
|
Gift From Ann S. Bowers '59 Creates New College of Computing and Information Science
Cornell Chronicle
Melanie Lefkowitz
December 17,
2020
A nine-figure gift from Cornell University alumnus (class of 1959) and philanthropist Ann S. Bowers will be used to build the Cornell Ann S. Bowers College of Computing and Information Science (CIS). The college will support growth in CIS, where half of all Cornell undergraduates take at least one course and enrollment is gaining at a rate faster than anywhere at the school. The new building that will house the new college also will support endowments for CIS faculty and students. Cornell's CIS program was one the first in the U.S. to integrate computer science and information science, focusing on technology's social impact. Cornell's Kavita Bala said Bowers' gift "will propel Cornell to lead the way in addressing the technological and societal challenges of our time."
|
UCI Researchers Create Model to Calculate Covid-19 Health Outcomes
UCI News
December 17, 2020
University of California, Irvine (UCI) health sciences researchers have developed a machine learning model to predict the likelihood of Covid-19 patients worsening within 72 hours, and requiring either a ventilator or intensive care unit (ICU) management. UCI's Daniel S. Chow said, "The goal is to give an earlier alert to clinicians to identify patients who may be vulnerable at the onset." The model is paired with a decision-making framework that uses patients' medical histories to help determine which patients will need critical care, and which can go home. The model used UCI Health patient data to create an algorithm that employs preexisting conditions, hospital test results, and demographic data to calculate the probability a patient will need a ventilator or ICU care. The model's predictions at UCI Health were found to be about 95% accurate.
|
Spanish Art Show Spotlights 'Hidden' Digital Divide in Pandemic
Reuters
Sophie Davies
December 18, 2020
An exhibition at Barcelona???s Analog Museum of Digital Inequality in Spain highlights a "hidden" digital divide in the Covid-19 pandemic, which disproportionately impacts women and low-income and ethnic minorities. The exhibition features a painting by Yaiza Ares that spotlights the gender gap by reinterpreting American realist Edward Hopper???s "Hotel Room." The work depicts a woman sitting on a bed and looking at text on an iPad that reads: "Only 17% of technology specialists in Europe are women." Also showcased is a pot by Maria Melero that embodies the digital divide's effect on children, picturing some children sitting studiously using computers, while others watch look on wistfully, apparently wishing they had their own screens. Isabella Longo, project director at BIT Habitat, the nonprofit behind the exhibition, said, ???Technological inequality is a hidden problem, (but) it has become especially obvious throughout this unprecedented year.???
|
AI Can Run Simulation Faithful to Physical Laws
Kobe University (Japan)
December 18, 2020
Researchers at Japan's Kobe University developed an artificial intelligence (AI) model that can simulate phenomena based on an unknown mechanism or formula, while preserving the laws of physics. The technology aims to enable predictions for phenomena that have been difficult to simulate, like wave motion or fracture mechanics, and speed up the simulations themselves. The researchers created a new digital version of backpropagation using automatic differentiation, which makes it possible to preserve physical laws in the digital world. The researchers also used geometric approaches to apply the technique to the prediction of a wider range of phenomena.
|
City Researchers Reveal Link Between Coding of Cryptocurrencies and Their Market Behavior
City University of London (U.K.)
John Stevenson
December 17, 2020
A connection between the coding of cryptocurrencies and their market behavior has been identified by researchers at the U.K.???s City University of London. The researchers studied 297 cryptocurrencies whose code is stored in GitHub, and whose average daily trading volume during their lifetime was higher than $100,000. They found that 4% of developers contribute to the code of two or more cryptocurrencies, raising questions about the transparency of the coding process. City University's Andrea Baronchelli said, "Cryptocurrencies are open source digital objects traded as financial assets that allow, at least theoretically, everyone to directly shape both an asset structure and its market behavior. Our study, identifying a simple event in the development space that anticipates a corresponding behavior in the market, establishes a first direct link between the realms of coding and trading."
|
iPhones Vulnerable to Hacking Tools for Months, Researchers Say
The Guardian (U.K.)
Alex Hern
December 20, 2020
Researchers at Canada's University of Toronto (U of T) said spyware sold by Israel???s NSO Group incorporated a zero-click zero-day exploit that could commandeer iPhones via a flaw in iMessage. They said the "Kismet" tool would leave no visible sign of implantation, could be installed by sending a message that victims did not have to click on, and worked even on phones running the latest iOS version. The researchers reported 37 known examples of Kismet being used by NSO Group clients against journalists covering news in and around the Middle East, yet "we suspect that the infections ??? were a minuscule fraction of the total attacks used with this exploit." The U of T team said logs from compromised phones suggested Kismet, or a similar exploit, has been in use for more than a year.
|
Making Smart Thermostats More Efficient
MIT News
December 18, 2020
Researchers from the Massachusetts Institute of Technology (MIT) and Russia's Skolkovo Institute of Science and Technology (Skoltech) collaborated on the design of a smart thermostat that uses data-efficient algorithms to learn optimal temperature thresholds within seven days. The thermostat's algorithms use reinforcement learning and are event-triggered, making decisions only in response to events defined by specific conditions reaching a threshold. Skoltech's Henni Ouerdane said, "A learning thermostat can potentially learn how to adjust its set-point temperatures in coordination with other HVAC devices, or based on its prediction of electricity tariffs in order to save energy and cost."
|
|
