Welcome to the September 25, 2017 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."
|
|
|
As More Women Enter STEM Fields, Difficulties Remain
The Wall Street Journal
Melissa Korn
September 25, 2017
A new report from researchers at Georgetown University suggests women may be more inclined to drop out of certain science, technology, engineering, and technology fields when they get low grades in disciplines that are stereotypically male and where men are already overrepresented. Consequently, the report notes it is significant when movement toward gender parity in such fields transpires. An analysis of U.S. Department of Education data found 10 schools awarded at least 100 computer science degrees to women at the bachelor's level and higher in the 2015-2016 academic year--up from only one in the 2011-2012 year. Another study estimated the University of Southern California awarded 848 bachelors and higher degrees in computer science in 2016. Concurrently, the school's percentage of women graduates rose from 17.8 percent to 29 percent. Meanwhile, Stanford University's share of female computer science degree recipients climbed from 14.3 percent to 25 percent between 2012 and 2016.
*May Require Paid Subscription
|
Internet Rulemaking Is Going to Get More Complicated
NextGov.com
Mohana Ravindranath
September 21, 2017
A new report from the Internet Society, a think tank co-founded by Google chief Internet evangelist and former ACM president Vint Cerf, who shared the 2004 ACM A.M. Turing Award with Robert E. Kahn, sees a strong need for "multistakeholder" governmental/societal collaboration on a "consensus policy" for the Internet. The report warns of cybersecurity measures increasingly undermining personal rights and freedoms through censorship and other means if "multilateral" government policies continue. "Measures that may be intended to secure cyberspace will increasingly undermine personal rights and freedoms," the report predicts. "If current trends are any indication, more and more governments will restrict and control Internet use and access through censorship, network shutdowns, and other means." The report notes the complex issues surrounding privacy and national security should drive governments to seek input from stakeholders outside government, while it also warns "weakening encryption technologies will create new vulnerabilities and cyberthreats."
|
Distrustful U.S. Allies Force Spy Agency to Back Down in Encryption Fight
Reuters
Joseph Menn
September 21, 2017
The U.S. National Security Agency (NSA) reportedly has been forced by an international coalition of cryptography experts to back off from pressing the independent International Organization for Standardization (ISO) to globally standardize several data encryption methods amid suspicion among U.S. allies. Academic and industry specialists from Germany, Japan, Israel, and elsewhere are concerned NSA was promoting the new techniques not because they were good encryption tools, but because it knew how to crack them. Following a series of closed-door meetings around the world over the past three years, which discussed whether ISO should approve two NSA data encryption techniques known as Simon and Speck, NSA has agreed to drop all but the most powerful versions of the techniques. Many experts who took part in the approval process for Simon and Speck were concerned NSA would gain a "back door" into coded transmissions if it were able to crack the encryption techniques.
|
NSF-Funded 'SLATE' Platform to Stitch Together Global Science Efforts
HPCwire
Rob Mitchum
September 21, 2017
The U.S. National Science Foundation's $4-million Services Layer At The Edge (SLATE) project is a multi-institutional effort to deliver technology that streamlines the linkage of university and laboratory data center capabilities to the national cyberinfrastructure ecosystem. Once set up, SLATE connects local research groups with remote collaborators so central research teams can automate the sharing of data, software, and computing tasks among institutions while sparing local system administrators from installing and operating customized scientific computing services. SLATE also aims to extend the presence of domain-specific "science gateways" and multi-site research platforms. SLATE employs best-of-breed data center virtualization elements and software-defined networking to facilitate automation of lifecycle management tasks by domain experts, easing the creation of scalable platforms that connect research teams, institutions, and resources, thus expediting science while lowering operational costs and development time. SLATE only needs commodity components, and can be used for distributed systems across all data center types and scales, supporting ubiquitous, science-motivated cyberinfrastructure.
|
National Day of Civic Hacking Events Aimed to Be More Inclusive
Government Technology
Zack Quaintance
September 21, 2017
Several of the events held at the fifth annual National Day of Civic Hacking on Saturday stressed the inclusion of more attendees who would normally regard themselves as non-technological experts. Code for Maine's Nick Kaufmann notes his group's Portland Civic Design Festival is an attempt to boost inclusivity, which he hopes will establish a bridge between civic hackers and people with related skills and viewpoints, such as urban planning innovators and architects. Meanwhile, organizers in Sacramento, CA, invited non-technologists to address the challenge of housing affordability, which is a major problem for California. Code for America director Erie Meyer calls the National Day of Civic Hacking a "culmination and leveling up of work that's been going on all year." She says, "Code for America is just one part of a huge universe of people who are working restlessly to try and make government work better, particularly for the most vulnerable people."
|
A Brain Built From Atomic Switches Can Learn
Quanta Magazine
Andreas von Bubnoff
September 20, 2017
Researchers at the University of California, Los Angeles are constructing a device the California NanoSystems Institute's Adam Stieg says is "inspired by the brain to generate the properties that enable the brain to do what it does." The device is a mesh of highly interconnected silver nanowires that is self-configured out of random chemical and electrical processes. This network contains 1 billion artificial synapses for each square centimeter, and experiments found it can execute simple learning and logic operations, as well as filtering out unwanted noise from received signals. Instead of using software, the researchers leverage the network's ability to distort an input signal in various ways, depending on where the output is quantified; this implies voice- or image-recognition applications. Another implication is the mesh could support reservoir computing, enabling users to select or mix outputs in such a manner that the result is a desired computation of the inputs.
|
Cornell, Carnegie Researchers Aim to Hold Computers Accountable for Their Decisions
Campus Technology
Joshua Bolkan
September 18, 2017
Researchers at Cornell and Carnegie Mellon universities will launch a U.S. National Science Foundation-funded project that seeks to make computers accountable for their decisions and respectful of privacy. One issue the team will probe is whether machine-learning systems leak private information about the datasets they are educated on via the conclusions at which they arrive. "This project will be a great opportunity to investigate the extent to which having access to the output of machine-learning systems reveals sensitive information and, in turn, how to improve machine learning to be more privacy friendly," says Cornell Tech professor Thomas Ristenpart. The researchers also will investigate the development of protective measures against the inclusion of real-world prejudices in computer applications. "A key innovation of the project is to automatically account for why an automated system with artificial intelligence components exhibits behavior that is problematic for privacy or fairness," notes CMU's Anupam Datta.
|
USC ISI Study Reveals How Information Spreads on Social Media
USC Viterbi School of Engineering
Caitlin Dawson
September 22, 2017
Researchers at the University of Southern California's Viterbi School of Engineering (USC Viterbi) have co-authored a study on how positive behavior is reinforced on social networks. In collaboration with the Technical University of Denmark, USC Viterbi professor Emilio Ferrara deployed a network of 39 social bots on Twitter coded to spread positive messages. Each bot automatically recorded when a target retweeted intervention-related content and also each exposure that had occurred prior to retweeting. "We found that bots can be used to run interventions on social media that trigger or foster good behaviors," Ferrara says. He also notes information is much more likely to spread virally when multiple sources expose people to the same piece of information multiple times, which he calls a "cumulative reinforcement effect." Ferrara says, "We have seen empirically that when you are exposed to a given piece of information multiple times, your chances of adopting this information increase every time."
|
Robot Judges? Edmonton Research Crafting Artificial Intelligence for Courts
CBC News (Canada)
Wallis Snowdon
September 19, 2017
University of Alberta professor Randy Goebel's team in Canada is working with researchers in Japan to develop artificial intelligence software for legal reasoning, after they successfully created an algorithm that can pass the Japanese bar exam. Goebel says the software is being designed to weigh contradicting legal evidence, rule on cases, and forecast the outcomes of future trials. "The next stage is to become more aggressive and [develop] not just yes-no questions, but do free-form questions," Goebel says. The new algorithm already has found use in the real world for studying legal precedents and outlining the best path to a legal victory. "A lot of the work that lawyers do in law offices is so-called discovery, trying to find out what data, what documents are related to a case that you're managing," Goebel says. "Speeding that up by having computer programs that do a more in-depth analysis of the language can save time and money."
|
Researchers Use Wikipedia to Give AI Context Clues
BYU News (UT)
Andrea Christensen
September 19, 2017
Researchers at Brigham Young University (BYU) are combining linear algebra and Wikipedia to give artificially intelligent (AI) agents contextual hints. "The common-sense understanding of what you can do with objects is utterly missing [in AI agents], and we end up with robots who will spend thousands of hours trying to eat the table," says BYU undergraduate Ben Murdoch. The team downloaded Wikipedia and processed it through a previously established algorithm that studies words in their contexts. The researchers then assessed their methodology by applying it to a series of text-based adventure games in which a player and agent engage in back-and-forth text interactions, with the agent offering a situation and the player responding with a written phrase. The team says its technique improved the computer's performance on 12 out of 16 games. The researchers' goal is to help build androids capable of navigating and interacting intelligently with their surroundings.
|
Real or Fake? Creating Fingers to Protect Identities
MSUToday
Jessi Adler
September 19, 2017
Michigan State University professor Anil Jain and a colleague have for the first time designed and created a fake finger containing multiple properties of human skin. Referred to as a spoof, this fake finger has been used to test two predominant types of fingerprint readers--optical and capacitive--to help determine their resilience to spoof attacks. The reproduced fingers were created using a combination of materials, including conductive silicone, silicone thinner, and pigments. "It will help motivate designers to build better fingerprint readers and develop robust spoof-detection algorithms," Jain says. He has begun work on designing and building a low-cost fingerprint reader to test spoof-detection capabilities, which Jain says could be built in a few hours by others in the fingerprint-recognition community to test for real versus fake fingerprints. Jain's lab also is working on algorithms that will make the reader more resilient to spoof-presentation attacks.
|
New Army Models Predict Number of Cyberattacks That Pierce Company Networks
U.S. Army Research Laboratory
September 19, 2017
A new study from the U.S. Army Research Laboratory suggests the number of cyber intrusions on a company or government organization's computer network can be predicted. The researchers propose four generalized linear models (GLMs) to predict the number of successful cyber intrusions into an organization's computer network. One model, the generalization of the Poisson regression model to the negative binomial GLM model, predicts the response variable notably better than others. The researchers note the models also demonstrate the intrusions data exhibit sufficient regularity, in a statistical sense, and the construction of a practically useful predictive model is feasible. In addition, they showed the number of violations of an organization's internal cybersecurity policies is a strong predictor of the number of intrusions. The team says a model of this nature may offer clues toward enhancing the security posture and perhaps the design and operation of an organization's computing systems and networks.
|
|
