Welcome to the August 20, 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."

robot and child Robots Have Power to Significantly Influence Children's Opinions
University of Plymouth (United Kingdom)
Alan Williams
August 15, 2018

Researchers at the University of Plymouth in the U.K. have demonstrated that young children are much more susceptible than adults to having their opinions and decisions influenced by robots. A comparison of this susceptibility between children and adults found children between seven and nine were more likely to give the same responses as humanoid robots, even if the robots were clearly wrong. Using the Asch paradigm, the researchers asked participants to look at a screen showing four lines and to identify which two were the same length. Solitary people almost never make a mistake but when performing the experiment with others, they tend to follow what others are saying. Solitary children scored 87% on the test, but when the robots joined in their score fell to 75%. Among the incorrect answers, 74% matched the robot's answers. The researchers say the study should inform policy for using autonomous social robots as aids for education professionals or child therapists.

Full Article

secure convolutional neural network, illustration More Efficient Security for Cloud-Based Machine Learning
MIT News
Rob Matheson
August 17, 2018

Massachusetts Institute of Technology researchers have developed an encryption method that secures data in online neural networks, without drastically slowing runtimes. The GAZELLE system combines homomorphic encryption and garbled circuits to bypass the drawbacks of both techniques. A user uploads ciphertext to a cloud-based convolutional neural network (CNN) while garbled circuits run on their own computer. The CNN performs all computation in the linear layer, and sends some data to the nonlinear layer so the CNN and user share the data; the user conducts some computation on garbled circuits, and sends the data back to the CNN. By dividing and sharing the workload, GAZELLE limits homomorphic encryption to complex math one layer at a time and restricts communication of the garbled circuits to only the nonlinear layers. The last step ensures both homomorphic and garbled circuit layers maintain a common randomization scheme: data is split into separate segments sent to separate parties, who synchronize their parts to reconstruct the full data.

Full Article
A Twist in Graphene Could Make for Tunable Electronic Devices
IEEE Spectrum
Dexter Johnson
August 16, 2018

Researchers from Columbia University, the National Institute for Materials Science in Japan, and the National Center for Scientific Research in France have demonstrated proof of principle for a twisting method using graphene/boron nitride heterostructures. They say the method allows control of the graphene's rotation and the dynamic variance of electrical, optical, and mechanical properties. The team thinks this could lead to new classes of electronic devices. Low friction between the graphene and other two-dimensional materials makes the twisting possible, and there is no strong chemical bonding between the crystal planes, enabling them to slide easily over one another. Columbia's Cory Dean says this property means devices could be intentionally designed to be rotatable. According to Dean, "A single material that can be locally 'twisted' to realize each of these components could enable significant new engineering opportunities."

Full Article

LiveTag tags with photographic paper at left LiveTag Is Out to Make Dumb Objects Smart
New Atlas
Ben Coxworth
August 16, 2018

University of California, San Diego researchers are developing a Wi-Fi-based system using inexpensive tags that can be stuck to everyday non-electronic objects. These "LiveTags" feature patterns of copper foil printed onto a flexible paper-like substrate without any other electronic parts. Each tag reflects radio signals that are being emitted by a Wi-Fi router, which are picked up by a receiver. As a demonstration, the researchers created a "paper-thin" interface that could be attached to any surface and used to remotely-control a music player. LiveTag can only sustain a communications range of one meter, and the team aims to extend that range and manufacture the tags via paper-and-ink printing.

Full Article

silicon germanium nanotransistor Transistor Technology May Improve Speed, Battery Life for Computers, Mobile Phones, Other Electronics
Purdue University News
August 14, 2018

Purdue University researchers have developed a technology design for field effect transistors that incorporates lasers, which could yield better switching behavior for computers and devices than traditional field effect transistors. The team says the blending of a quantum cascade laser and transistor technologies into one design concept will help integrated circuit manufacturers aiming to build smaller and more transistors per unit area. The technology is designed to boost the speed, sensitivity, and battery life of computers, mobile phones, and other digital devices. The device boasts a large on-current and a low off-current with a small subthreshold swing, enabling greater speed and energy savings. Combining or stacking several switching mechanisms that simultaneously turn the transistor on or off is key, and the team is striving to optimize the technology and the effectiveness of the design.

Full Article
Networking Vendors Patch Against New Cryptographic Attack
Help Net Security
Zeljka Zorz
August 15, 2018

The Internet protocol IPSec is vulnerable to attack because the Internet key exchange (IKE) protocol can allow attackers to retrieve session keys and decrypt connections, according to researchers from Ruhr-Universitat Bochum in Germany and the University of Opole in Poland. "[Reusing] a key pair across different versions and modes of IKE can lead to cross-protocol authentication bypasses, enabling the impersonation of a victim host or network by attackers," the researchers say. They exploited a Bleichenbacher oracle in an IKEv1 mode, where RSA encrypted nonces are used for authentication. The researchers broke the RSA encryption-based modes and RSA signature-based authentication in both IKEv1 and IKEv2. An offline dictionary attack can also occur in PSK (Pre-Shared Key) -based IKE modes, covering all available authentication mechanisms of IKE. Networking vendors say they have pushed out fixes for the vulnerability, or removed the vulnerable authentication method from their devices’ firmware.

Full Article
U of T Experts in AI Explore a Classical Problem of Computer Science
U of T News
Nina Haikara
August 14, 2018

Researchers at the University of Toronto (U of T) in Canada are exploring synergies between fast, effective algorithms in artificial intelligence for automated planning and program synthesis, or generating a computer program automatically from a specification of user intent. U of T's Alberto Camacho says he used linear temporal logic (LTL) to develop a practical tool for synthesis, expressing the user's intent in instructions similar to English-language instructions. "I can say something to you, and if you misunderstand, we can have a conversation back and forth [to clarify]," says U of T's Sheila McIlraith. "But when we care about synthesizing safety-critical systems...it really matters that the system understands what we're asking it to do." Camacho says the team is applying automated planning algorithms to LTL synthesis. The program that is produced is correct by construction, and every run of that program satisfies the LTL statements with which it is provided.

Full Article

An image of two empty speech bubbles. Researchers Have Figured Out Ways to Dodge Censorship on WeChat
Echo Huang
August 15, 2018

China-based WeChat users are creatively bypassing censorship with techniques that include sharing images instead of text, according to an investigation by the Citizen Lab at the University of Toronto in Canada. The researchers concentrated on Tencent's WeChat Moments to identify two image-filtering algorithms used in the application. The first program employs optical-character recognition to detect sensitive text in images, while the other compares an uploaded picture against blacklisted images. The team determined censorship could be circumvented by mirroring or rotating the image, altering its aspect ratio, blurring the photo, adding a sufficiently large border to the smallest dimensions of an image, or adding a large border to the largest dimensions of an image as well as adding a sufficiently complex pattern to it. The researchers say Tencent appears to be maintaining a roster of blacklisted images that government officials can update.

Full Article

Matter and antimatter in the nanoscale magnetic universe Magnetic Antiparticles Offer New Horizons for Information Technologies
Uppsala University
August 15, 2018

Researchers at Uppsala University in Sweden, the Universities of Kiel and Mainz in Germany, and University Paris-Saclay in France have discovered behavior involving the antiparticle equivalent of skyrmions in a nanometer-thick ferromagnetic material, using computer simulations that can precisely model the materials' magnetic properties. Uppsala's Ulrike Ritzmann says the work "suggests that antiskyrmions could be a ready source of skyrmions," which would be critical for applications that use skyrmions to transmit and store information. The simulations revealed that gradually raising the applied current causes skyrmions to travel in straight lines, while antiskyrmions experience curved trajectories resembling trochoids. Moreover, boosting the energy transferred to the system from the currents causes this motion to transition to periodic generation of skyrmion-antiskyrmion pairs, in which each antiskyrmion created subsequently becomes a new source of pairs.

Full Article

Photo of AI4All alumni Aarzu Gupta and Lili Sun, who developed a drone program that can identify wildfires before they spread too far. AI Has a Racial Bias Problem. Google Is Funding Summer Camps to Try to Change That
USA Today
Ashley Wong
August 10, 2018

The Oakland, CA-based AI4All nonprofit will use a $1-million grant from Google.org to broaden its outreach to young underrepresented minorities and women, as one avenue of addressing the problem of racial bias in artificial intelligence (AI). AI4All is based on a two-week summer camp program started by Stanford University, targeting high school students who are women, people of color, or from low-income families. The chief purpose of the grant is to establish a free, online AI curriculum course that is accessible to anyone in the world. Google.org's Hannah Peter says, "We really need for AI to be made by diverse creators, and that starts with people having access to the learning opportunities to understand at its core what AI is and how it can be applied."

Full Article
Quantum Material Is Promising 'Ion Conductor' for Research, New Technologies
Purdue University News
August 15, 2018

Purdue University researchers have demonstrated a method for shuttling lithium ions back and forth into the crystal structure of the quantum material samarium nickelate. They say the method could find potential use in batteries, "smart windows," and brain-inspired computers with artificial synapses. The team found doping the material with lithium ions causes the crystal to expand and increases ion conduction, an effect that also extends to other types of ions, especially sodium ions. Applying a voltage triggered the ions' occupation of spaces between atoms in the crystal lattice, which could yield a more efficient method for storing and conducting electricity. In addition, the ions stayed in place after the current was deactivated, a "non-volatile" behavior that might be tapped for computer memory.

Full Article

Colorful image of cyber algorithms Crowdsourcing Algorithms to Predict Epileptic Seizures
Melbourne Newsroom
August 9, 2018

More than 10,000 crowdsourced algorithms can enable clinically relevant epileptic seizure prediction across a broad spectrum of patients, according to a study from the University of Melbourne in Australia. The algorithms were developed via a contest hosted on the Kaggle.com data science competition platform, and based on long-term electrical brain activity recordings acquired in 2013 from the world-first clinical trial of the implantable NeuroVista Seizure Advisory System. Melbourne's Levin Kulhmann says the competition attracted more than 646 participants and 478 teams worldwide. The algorithms can differentiate 10-minute inter-seizure from pre-seizure data clips, with the leading programs tested on patients with the lowest seizure prediction performance based on previous studies. Kuhlmann says the assessments "revealed on average a 90% improvement in seizure prediction performance, compared to previous results."

Full Article
Web Security Gets a Boost as TLS Gets Major Overhaul
Liam Tung
August 14, 2018

The Internet Engineering Task Force (IETF) has finalized Transport Layer Security version 1.3 (TLS 1.3), which will become the chief protocol for securing Web communications on HTTPS sites and applications. The previous version can be deployed securely, but several high-profile vulnerabilities have exploited optional parts of the protocol and outdated algorithms, according to IETF engineers, who say TLS 1.3 eliminates many of these problematic options. Data from Mozilla shows about 5% of Firefox connections are TLS 1.3, while Facebook says more than 50% of its Internet traffic is secured with TLS 1.3. IETF's working group chairs say, "If you manage a website or other online service, the servers and infrastructure you use are likely to start using TLS 1.3, though it is worth double-checking with your providers."

Full Article
ACM Transactions on Internet of Things (TIOT)
ACM Publications

Association for Computing Machinery

2 Penn Plaza, Suite 701
New York, NY 10121-0701

ACM Media Sales

If you are interested in advertising in ACM TechNews or other ACM publications, please contact ACM Media Sales or (212) 626-0686, or visit ACM Media for more information.

To submit feedback about ACM TechNews, contact: [email protected]