Welcome to the January 17, 2018 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."
|
|
|
Alibaba and Microsoft AI Beat Humans in Stanford Reading Test
Financial Times
Louise Lucas
January 15, 2018
Artificial intelligence (AI) programs from China and the U.S. last week outperformed humans in a global reading comprehension test. The algorithms of China-based Alibaba and Microsoft tied for first place, achieving accuracy levels a few basis points above humans' 82.3 percent in providing exact matches to questions from the Stanford Question Answering Dataset. Alibaba's Luo Si on Monday called his company's AI victory "a milestone," and noted the technology has multiple uses ranging from customer service to museum tutorials to medical queries, some of which are already being managed by chatbots worldwide. Both AIs' triumph on the reading test symbolizes the aggressive technological competition between China and the U.S., with China planning to leverage its massive datasets compiled from its populace and deep governmental investments to surpass the U.S. in cultivating a $150-billion industry.
*May Require Free Registration
|
No Evidence to Support Link Between Violent Video Games and Behavior
University of York
Samantha Martin
January 16, 2018
Researchers at the University of York in the U.K. have conducted experiments involving more than 3,000 participants, demonstrating that video game concepts do not "prime" players to behave in certain ways and that increasing the realism of violent video games does not necessarily boost aggression in gamers. Previous studies on the priming effect provided mixed conclusions; the University of York researchers expanded the number of participants and compared different types of gaming realism. In one study, participants played a game where they had to either be a car avoiding collisions with a truck, or a mouse avoiding a cat. Following the game, the players were shown various images and asked to label them as either a vehicle or an animal. "We found that the priming of violent concepts, as measured by how many violent concepts appeared in the word fragment completion task, was not detectable," says the University of York's David Zendle.
|
NSF Begins Awarding STEM Ed Grants for Hispanic-Serving Institutions
Campus Technology
Joshua Bolkan
January 16, 2018
The U.S. National Science Foundation (NSF) has started issuing its first awards via its Improving Undergraduate STEM Education: Hispanic Serving Institutions (HSIs) program, which is designed to improve science, technology, engineering, and math (STEM) education, build capacity, and improve graduation rates. This year's awards will go to seven conferences that worked to gather stakeholder input on STEM education at HSIs. The NSF program is now accepting grant proposals in the areas of building capacity and HSIs that are new to the agency. The grants will support projects in the areas of critical transitions, which will encompass projects designed to help Hispanic students through important education transitions; cross-sector partnerships, which will include projects aimed at developing partnerships that build faculty capacity and offer student opportunities in STEM research and education; and research on broadening participation in STEM, which will focus on projects to improve the retention and graduation rates of students pursuing STEM degrees at HSIs.
|
Smart Buildings That Can Manage Our Electricity Needs
Swiss Federal Institute of Technology in Lausanne
Clara Marc
January 15, 2018
Researchers at the Swiss Federal Institute of Technology of Lausanne (EPFL) have developed a system to collect data on consumers' energy usage and comfort within buildings. They say the system gathers data sent from connected devices to create an overall perspective of a building's electricity needs over time and by room, and that data is then sent to a smart grid to predict energy needs and decide how best to allocate resources. The researchers say the new interface will be able to combine disparate datasets and cover various energy needs; they also note it is highly flexible and ensures the best possible data protection. The team designed the system to be as generic as possible, including the use of open source computer code. "Our researchers are working on how to improve energy metering, communications between connected devices, and data collection and use, in order to enhance coordination between buildings and the grid," says EPFL's George Lilis.
|
UMass Center for Data Science Partners With Chan Zuckerberg Initiative to Accelerate Science and Medicine
UMass Amherst News
January 16, 2018
The University of Massachusetts Amherst's (UMass Amherst) Center for Data Science has announced a partnership with the Chan Zuckerberg Initiative (CZI) to accelerate science and medicine via the Computable Knowledge project. The project's purpose is to develop an intelligent and navigable map of scientific knowledge using knowledge representation and reasoning. Upon the project's completion, the service will be accessible via CZI's free Meta platform and will help researchers track important discoveries, find patterns, and provide insights among an up-to-date collection of published scientific texts, including more than 60 million articles. "We believe the result will be a first-of-its-kind guide for every scientist, just as map apps are now indispensable tools for navigating the physical world," says UMass Amherst professor Andrew McCallum. "We hope our results will help solve the mounting problem of scientific knowledge complexity, democratize scientific knowledge, and put powerful reasoning in the hands of individual scientists."
|
UVA Engineering Tapped to Lead $27.5-Million Center to Reinvent Computing
HPCwire
January 16, 2018
The University of Virginia (UVA) School of Engineering & Applied Science will use a $27.5-million grant to establish the national Center for Research in Intelligent Storage and Processing in Memory (CRISP), where researchers from eight universities will work to eliminate the partition between memories that store data and processors that operate on the data. This separation has led to what scientists call the "memory wall," in which data access significantly slows down performance. "Solving these challenges and enabling the next generation of data-intensive applications requires computing to be embedded in and around the data, creating 'intelligent' memory and storage architectures that do as much of the computing as possible as close to the bits as possible," says UVA professor Kevin Skadron. CRISP researchers will build processing capabilities within memory storage, couple processors with memory chips in three-dimensional stacks, and study how other facets of computer systems will have to change when computer architecture is rethought.
|
Data, Algorithms, and Fairness Panel
CCC Blog
Helen Wright
January 11, 2018
The Data, Algorithms, and Fairness panel at the Computing Community Consortium Symposium on Computing Research: Addressing National Priorities and Societal Needs focused on how data-driven and algorithmic decision-making increasingly shapes how businesses target ads to consumers, how police departments track individuals or groups, how banks select loan applicants, how employers hire, and how colleges and universities make admissions and financial aid decisions. As data-driven decisions increasingly impact all aspects of daily life, there is a pressing need to ensure they do not become tools for promoting prejudice and inequality, along with a growing emphasis on mitigating algorithmic bias and boosting algorithmic transparency. Panelists suggested non-empirically trained people writing about such issues should consult with and learn from individuals in the field, as well as compare algorithmic bias to what exists in current decision-making and challenge industry definitions of fairness. They noted there are different definitions of fairness than industry definitions.
|
Nanostructured Gate Dielectric Boosts Stability of Organic Thin-Film Transistors
Georgia Tech News Center
John Toon
January 12, 2018
Researchers at the Georgia Institute of Technology (Georgia Tech) say they have developed a nanostructured gate dielectric that could address a major hurdle to expanding the use of organic semiconductors for thin-film transistors. The new structure, consisting of a fluoropolymer layer and a nanolaminate made from two metal oxide materials, serves as a gate dielectric, protecting the organic semiconductor, and enabling the transistors to operate with increased stability. "This could be the tipping point for organic thin-film transistors, addressing long-standing concerns about the stability of organic-based printable devices," says Georgia Tech professor Bernard Kippelen. He notes the architecture uses alternating layers of aluminum oxide and hafnium oxide--five layers of one followed by five layers of the other, repeated 30 times on top of the fluoropolymer--to produce the dielectric. The oxide layers are produced with atomic layer deposition, and the nanolaminate is basically immune to the effects of humidity.
|
'Earworm Melodies With Strange Aspects'--What Happens When AI Makes Music
Horizon Magazine
Kevin Casey
January 12, 2018
The FlowMachines project, funded by the European Union's European Research Council, has produced "Hello World," the first entire studio album co-created by artists and artificial intelligence (AI). The album was produced by 15 artists, music producer Benoit Carre, and software designed by computer scientist Francois Pachet. Pachet says the software uses neural networks that learn from experience by forming connections over time, mimicking the biological networks of the human brain. He notes a musician initially provides "inspiration" to the software by exposing it to a collection of songs, and after the system understands the given style, it creates a new composition. "The system...analyzes the music in terms of beats, melody, and harmony," Pachet says, "and then outputs an original piece of music based on that style." The researchers note they had to make sure the software could adapt to the creative workflow of musicians without becoming annoying.
|
U.S. Army Scientists Improve Human-Agent Teaming by Making AI Agents More Transparent
ARL News
January 11, 2018
Researchers at the U.S. Army Research Laboratory (ARL) have developed ways to improve collaboration between humans and artificially intelligent agents. The researchers say they have enhanced agent transparency, which refers to a robot, unmanned vehicle, or software agent's ability to communicate to humans its intent, performance, future plans, and reasoning process. In 2016, the U.S. Defense Science Board identified six barriers to human trust in autonomous systems, including low observability, predictability, directability, and low mutual understanding of common goals. The ARL researchers addressed these issues by developing the Situation awareness-based Agent Transparency (SAT) model, and measured its efficacy on human-agent team performance in human factors studies. One project, IMPACT, focused on examining the effects of levels of agent transparency, based on the SAT model, on human operators' decision-making. Meanwhile, the Autonomous Squad Member project involved a small ground robot interacting and communicating with an infantry squad.
|
Pulses of Light to Encrypt Data and Protect Security of Cryptocurrencies
USC Viterbi News
January 11, 2018
Researchers at the University of Southern California's Viterbi School of Engineering have developed a frequency comb they think could encrypt data and bolster the security of cryptocurrency. The team says the comb boosts the potential application of lasers by converting a single wavelength into multiple wavelengths. Not only is the tool infinitesimally smaller than traditional frequency combs, it also requires 1,000 times less power to operate, permitting mobile applications. The researchers deposited a single layer of a 25-atom organic molecule on the surface of a laser, eliminating the introduction of silicon. The team says frequency comb generators can be viewed as entangled photon generators, which is an important step toward quantum encryption. The researchers also note key technical challenges to overcome included shrinking the size and power requirements of the frequency comb, although they say there are still many integration and manufacturing challenges remaining before quantum cryptography on portable platforms can become routine.
|
Department of Energy Joins Quest to Develop Quantum Computers
Science
Adrian Cho
January 10, 2018
The U.S. Department of Energy (DoE) says it will participate in the development of quantum computers with a $40-million project to create practical quantum computing algorithms for chemistry, materials science, nuclear physics, and particle physics research. Among the challenges researchers believe quantum systems can help tackle are those that involve modeling innately quantum mechanical processes. Quantum computers also could assist in the design of materials from their atomic components on up. The bulk of the funding for the DoE initiative will go to the department's national laboratories, and University of Maryland in College Park professor Christopher Monroe thinks DoE scientists can take a lead development role. The DoE says the project is partly a response to the accelerating progress of commercial entities such as Google in the race to achieve quantum supremacy over classical computer systems on a test problem.
|
Q&A With Computer Science Professor Prasad Jayanti
The Dartmouth
Charles Chen
January 16, 2018
In an interview, Dartmouth College professor Prasad Jayanti discusses his specialization in concurrent algorithms, which he describes as a scenario in which "you have a problem and not one, but several processors--several computers--are coordinating with each other to help solve the problem." Jayanti notes this leads to more problems, and overcoming those problems is what the field of distributed computing is designed to accomplish. Among his current research areas is addressing the challenge of solving specific problems correctly and efficiently even when a system crashes and restarts. Jayanti suggests aspiring computer science students disabuse themselves of the myth of computer science as a solitary and isolating pursuit. He also observes that computer science is "intellectually fascinating. It challenges your mind; when you pass [the course], it pays you well and takes care of you more than adequately and it's open to everyone."
|
|
