Welcome to the February 3, 2023, 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."
|
|
|
Japanese Universities Rush to Set Up Data Science Departments
Nikkei Asia
February 2, 2023
Japan's Education Ministry has confirmed that 17 Japanese universities will launch data science and information technology departments at the beginning of the new school year in April, aimed at training as many as 1,900 students in order to ramp up digitization in businesses and government. Government estimates in 2019 projected up to 790,000 unfilled positions in 2030. Hitotsubashi University's forthcoming data science department will include artificial intelligence experts on staff and will educate students on data science applications in social science. The Ministry of Education hopes graduates of data science programs will have the skills needed to lead corporate digitization.
|
Using Robots to Fight Toxic Algae
USC Viterbi School of Engineering
Lillian Goodwin
February 2, 2023
A method devised by computer scientists and biologists at the University of Southern California can enable autonomous robots to identify the best sampling sites for toxic algae. The robots function as a "pre-survey" to check regions ahead of biologists, using a planning framework to search for traces of algae blooms while exploring bodies of water based on the biologists' preferences. An algorithm uses collected data to intelligently focus on areas to quantify toxic algae levels. The autonomous system model is "smarter" than the devices biologists typically use for the purpose because it takes advantage of an environmental simulation using informative path planning to ascertain the optimal route for fulfilling its objectives.
|
UAE Lunar Rover to Test First AI on the Moon with Canada
Space.com
Elizabeth Howell
January 28, 2023
A machine learning system developed by the Canadian space technology company Mission Control Space Services (MCSS) was the first artificial intelligence (AI) to reach beyond low Earth orbit. It was launched on a SpaceX mission Dec. 11 as a part of Japan's ispace lander to inform the decision-making of the UAE's Rashid rover as it searches the surface of the moon for minerals and other items. MCSS's algorithm will classify each pixel in the rover's navigation images, sent via the Japanese lander, by type of terrain. MCSS's Ewan Reid said, "That output will then be sent to the ground and will be used by scientists and engineers at our office in Ottawa, as well as at other Canadian universities, to help decide where the rover should go."
|
Child Welfare Algorithm Faces Justice Dept. Scrutiny
Associated Press
Sally Ho; Garance Burke
January 31, 2023
An algorithm used by a child protective services agency in the Pittsburgh area is drawing scrutiny from the U.S. Department of Justice's Civil Rights Division. The Allegheny Family Screening Tool is used by Pennsylvania’s Allegheny County Department of Human Services to assess a family's risk level to help determine whether social workers should investigate them. It has generated civil rights complaints over concerns it is biased against families with disabilities, including those with mental health issues. The tool uses personal data such as substance abuse and mental health records, as well as data related to Supplemental Security Income, to predict the risk of a child being placed in foster care within two years of a family investigation.
|
Voice-Activated System for Hands-Free DNA Handling
American Chemical Society
February 1, 2023
The American Chemical Society reports a new voice-activated, hands-free device offers a safer way to extract and pretreat bacterial DNA during disease outbreaks, and could assist scientists with disabilities in conducting research. The palm-sized device is comprised of a microfluidic chip with multiple chambers connected to six three-way solenoid valves operated by a micro-controller, and a voice-controlled smartphone app. The app sends an initiation signal to the micro-controller after an operation command is spoken, triggering the micro-controller to perform a series of steps. The micro-controller can load a sample, then wash and release purified DNA into a collection chamber. While the researchers found device performance not on par with traditional DNA extraction kits, they said it offers greater safety and convenience.
|
Wearable Sensor Provides Cardiac Imaging on the Go
UC San Diego Today
Emerson Dameron; Ioana Patringenaru
January 25, 2023
A team led by researchers at the University of California, San Diego (UC San Diego) has developed a portable ultrasound device that uses custom algorithms to measure how much blood is being pumped by the heart. The postage-stamp-sized patch captures images of the heart at different angles using ultrasound. The technology performs real-time analysis of a subset of the images, which have high spatial resolution, temporal resolution, and contrast. UC San Diego's Ruixiang Qi said the device provides "accurate and continuous waveforms of key cardiac indices in different physical states, including static and after exercise, which has never been achieved before."
|
Virtual Birkin Bags on Trial in Hermès Case Testing IP Rights
The Wall Street Journal
Erin Mulvaney
January 29, 2023
French luxury brand Hermès is suing digital artist Mason Rothchild in a New York court to prevent him from selling nonfungible tokens (NFTs) of its Birkin handbags, claiming it violates trademark law and dilutes its brand. Rothchild argues the First Amendment protects his "MetaBirkins" as artwork, while legal analysts say the case is an early test of exercising rights against unauthorized virtual assets. In his legal declaration, the artist maintains the NFTs are imaginary bags rather than replicas, intended to artistically explore conspicuous consumption. Hermès responded in court documents that Rothchild "seeks to make his fortune by swapping out Hermès' 'real-life' rights for 'virtual rights'.”
*May Require Paid Registration
|
Smart Contact Lens Diagnoses, Treats Glaucoma
Pohang University of Science and Technology (South Korea) Research Highlights
January 31, 2023
A smart contact lens developed by researchers at South Korea's Pohang University of Science and Technology can monitor intraocular pressure (IOP) in glaucoma patients and deliver medications in response to IOP levels. The smart contact lens integrates an IOP sensor that incorporates a flexible drug delivery system with gold nanotubes, a wireless power and communication system, and an application-specific integrated circuit (ASIC) chip. In tests on rabbits with glaucoma, the researchers found the device could measure IOP in real time and deliver timolol on demand in a dosage that matched the degree of IOP detected.
|
To Know Where the Birds Are Going, Researchers Turn to Citizen Science, Machine Learning
University of Massachusetts Amherst
February 1, 2023
Researchers at the University of Massachusetts Amherst and Cornell University have developed a predictive model that can forecast the destination of bird migration. BirdFlow uses data from Cornell's eBird Status & Trends database, which contains data on over 200 million annual bird sightings submitted by birders worldwide. The model runs that data through a probabilistic machine learning model, which has learned to predict the movement of individual birds using real-time GPS and satellite tracking data. In tests on 11 species of North American birds, the researchers found BirdFlow outperformed other bird migration tracking models and can make accurate migration-flow predictions without the use of real-time GPS and satellite tracking data.
|
NCSES Releases Report on Diversity in STEM
HPCwire
January 30, 2023
The latest analysis from the U.S. National Science Foundation's National Center for Science and Engineering Statistics on diversity in science, technology, engineering, and math (STEM) employment and education indicates increasing workforce representation of women and people of color in the past decade. Yet those groups, as well as disabled persons, remain underrepresented in STEM fields compared to the general population, according to the report. Although women and Hispanics in particular were found to have made significant inroads in STEM employment and education, those trends were not found to be consistent across all STEM occupations and disciplines. For example, minorities comprise a third of the STEM workforce that do not require college degrees, according to the study, in jobs that tend to be the lowest-paid and have the highest unemployment.
|
GPS Tracking, Simulations Show Optimal Locations to Help Bighorn Sheep Cross Freeways
Oregon State University News
Steve Lundeberg
January 30, 2023
A team of researchers at Oregon State University (OSU) and the U.S. National Park Service produced models that use global positioning system (GPS) data to help determine places in Southern California where desert bighorn sheep can safely cross interstate highways via overpasses. The researchers analyzed the movements of members of nine bighorn populations in the Mojave Desert using collated GPS data. This enabled them to identify landscape features the sheep prefer migrating through, and those they would rather avoid. The researchers modeled a cumulative 8,200 years in the lives of bighorn sheep on landscapes with and without freeways, and found they strongly correlate with suggested movement corridors derived from genetics data and roadkill statistics, offering a finer perspective on the benefits of a barrier-free artificial environment.
|
Stable Diffusion 'Memorizes' Some Images, Sparking Privacy Concerns
Ars Technica
Benj Edwards
February 1, 2023
An international team of artificial intelligence (AI) researchers has formulated an adversarial attack that can exfiltrate a small number of training images from latent diffusion AI image synthesis models such as Stable Diffusion. The researchers estimated an approximately 0.03% memorization rate out of 350,000 high-probability images from the Stable Diffusion training dataset. They also pointed out that this "memorization" is approximate, because the AI model cannot generate identical byte-for-byte duplicates of the training images. One AI authority suggested this research could impact potential image synthesis regulations if the AI models are designated "lossy databases" that can replicate training data.
|
Matching Algorithms Work Fine — Until Bypassed by External Links
UCLA Anderson Review
Monika Brown
February 1, 2023
Researchers at the University of California, Los Angeles, Stanford University, and Yale University unveiled a framework that enables matching algorithms to more efficiently maximize matches by integrating external and internal traffic. Partnering with the VolunteerMatch platform for linking nonprofits with volunteers, the researchers designed the Adaptive Capacity algorithm to maximize sign-ups in the presence of external traffic, based on an algorithm engineered to keep the percentage of remaining capacity the same for each resource. Simulations showed Adaptive Capacity outperformed VolunteerMatch's current recommendation algorithm. The researchers also identified a similarly performing but simpler algorithm that requires less data.
|
|
