Association for Computing Machinery
Welcome to the February 1, 2016 edition of ACM TechNews, providing timely information for IT professionals three times a week.

Updated versions of the ACM TechNews mobile apps are available for Android phones and tablets (click here) and for iPhones (click here) and iPads (click here).


New Technologies Give Government Ample Means to Track Suspects, Study Finds
The New York Times (01/31/16) David E. Sanger

U.S. intelligence agencies' persistent warning that encrypted communications will prevent them from tracking criminals is greatly distorted, as new technologies are creating many opportunities for the government to monitor suspects, according to a new study from Harvard University's Berkman Center for Internet and Society. The researchers say technologies are "being packed with sensors and wireless connectivity," which are expected to become the subject of court orders and subpoenas, and already are the target of the U.S. National Security Agency as it augments networks worldwide to track communications abroad via "implants." The Harvard study involved officials, technical experts, and others who are or have been on the forefront of counterterrorism. Harvard professor Jonathan Zittrain says the group was convened to discuss the issue and go beyond its sticking points "in part by thinking of a larger picture, specifically in the unexpected ways that surveillance might be attempted." He points out the current debate rarely acknowledges the Internet of Things, in which telemetry from everyday appliances could be obtainable via subpoena from governments around the world. "Law enforcement or intelligence agencies may start to seek orders compelling Samsung, Google, Mattel, Nest, or vendors of other networked devices to push an update or flip a digital switch to intercept the ambient communications of a target," the report warned.
View Full Article - May Require Free Registration | Return to Headlines | Share Facebook  LinkedIn  Twitter 

Obama Wants $4B for More Computer Science Education
IDG News Service (01/30/16) Blair Hanley Frank

U.S. President Barack Obama's proposed fiscal 2016 budget will include $4 billion to provide states with money to develop computer science programs, and another $100 million in grants available directly to school districts to further computer learning. In addition to encouraging the creation of computer science programs, the money will be used to build programs that attract students who typically are not served by that education segment. Regardless of whether or not Congress supports the plan, federal agencies will tap some of their existing funds to back similar programs. The U.S. National Science Foundation will invest $120 million in developing instructional materials for schools and resources to ensure teachers are prepared to educate students in computer science. Meanwhile, the U.S. Corporation for National and Community Service is pledging $17 million to support similar educator training. Obama's agenda includes improving U.S. citizens' computer science knowledge so they can participate more effectively in an increasingly computer-driven economy. "This isn't just a tech issue, this isn't just an education issue, computer science education is now an economic and social imperative for the next generation of American students," says Microsoft President Brad Smith.

Researchers Discover Malicious Computer Code
Ben-Gurion University of the Negev (Israel) (01/27/16) Andrew Lavin

Ben-Gurion University of the Negev (BGU) researchers say they have discovered and traced six botnets by analyzing data collected from past cyberattacks. The researchers, led by BGU professors Bracha Shapira and Lior Rokach, analyzed data captured by a "honeypot" network, and developed and implemented algorithms to identify the botnet by finding similar attack patterns that could be traced to their administrator. The researchers identified six botnets, each capable of inflicting significant criminal and monetary damage. The researchers also were able to identify whether the attack came from a real person or from a robot, and predict future attacks. The research was conducted at Deutsche Telekom (DT) Innovation [email protected] "This is the first time such a comprehensive study has been carried out and returned with unique findings," says DT [email protected] chief technology officer Dudu Mimran. The research was announced at CyberTech 2016 in Tel Aviv, Israel.

Microsoft Plumbs Ocean's Depths to Test Underwater Data Center
The New York Times (01/31/16) John Markoff

A solution to the expensive requirement of cooling data-center servers could be found underwater, as Microsoft researchers have tested a prototype data center that can function hundreds of feet below the surface of the ocean. The concept, called Project Natick, also may address computing's expanding energy demands because Microsoft is weighing combining the system either with a turbine or a tidal energy system to produce electricity. The initiative might lead to strands of giant steel tubes placed on the seafloor and connected via fiber optics, or to capsules suspended beneath the surface, with turbines driven by the ocean current. The researchers think mass production of the capsules could dramatically accelerate data-center deployment to only 90 days, realizing significant cost savings. Web services also could operate faster via this scheme thanks to lower latency. Microsoft's 105-day trial of a steel capsule placed 30 feet underwater near the California coast and controlled from offices at the Microsoft campus outdid expectations, as anticipated hardware failures and leaks did not materialize. The researchers currently are designing a new underwater system triple the size of the original prototype, to be built by the developer of an ocean-based alternative energy system.
View Full Article - May Require Free Registration | Return to Headlines | Share Facebook  LinkedIn  Twitter 

Graduate Students Work to Help Make Voting Process Easier
The Shorthorn (01/27/16) Brittany Harborth

University of Texas at Arlington (UTA) professor Chengkai Li and his team of graduate students are working to enable anyone to rapidly fact-check politicians' claims. The team is developing a problem-solving program called ClaimBuster, which learns from training data, according to Li. People participate in data collection by filling out surveys of what they deem worthy of fact-checking and importance. The software analyzes closed captions from political debates and ranks each claim by what has a higher likeliness of needing to be checked. Sentences are scaled by how check-worthy they are by color and number, leaving fact checkers to do the rest, notes Naeemul Hassan, research and development leader for the project. Number amounts and changes in verb tenses tend to lead to a higher check-worthiness score, Li notes. UTA professor Mark Tremayne is the co-principle investigator for the project, and discussed using the presidential debates for the project from the beginning, Li says. The team plans to display a more refined prototype website by April, which will have additional capabilities and features, according to Hassan.

Scientist Michelle Simmons Leading the Race to Build World's Fastest Quantum Computer
Australian Financial Review (01/28/16) Tim Dodd

Michelle Simmons at the University of New South Wales is regarded as a leading researcher in the development of a practical, programmable quantum computer. Simmons and her team last year successfully built a pair of qubits from two phosphorous atoms placed in a single silicon chip. They succeeded in "entangling" the atoms, so what happens in one atom is intrinsically linked to the other. The key goal of Simmons' research group, the Center for Quantum Computation and Communication Technology, is to create a quantum computer that works at unprecedented speed yet can be commercialized. Simmons points to studies estimating the development of quantum computers would benefit about 40 percent of a modern economy. Simmons and her team are now testing designs for three- and four-qubit systems, with each qubit made from a single phosphorous atom encased in one silicon chip. In December, Australia's government and other entities committed $46 million to creating a 10-qubit system, which the researchers hope to achieve by 2020. Simmons says about 300 qubits are required to build a quantum computer that provides a transformational boost in speed.
View Full Article - May Require Free Registration | Return to Headlines | Share Facebook  LinkedIn  Twitter 

New Chip Fabrication Approach
MIT News (01/27/16) Larry Hardesty

Massachusetts Institute of Technology (MIT) researchers and their colleagues say they have created the first chip-fabrication technique to enable different materials to be deposited in the same layer. The layers in the researchers' experimental chip are extremely thin, between one and three atoms thick. The researchers say the work could facilitate efforts to manufacture thin, flexible, and transparent computing devices that could be laminated onto other materials. "This offers us tremendous potential with numerous candidate materials for ultrathin circuit design," says MIT postdoctoral researcher Xi Ling. The technique has implications for the development of ultralow-power, high-speed computing devices known as tunneling transistors, and for the integration of optical components into computer chips. To assemble their laterally integrated circuits, the researchers first deposited a layer of graphene on a silicon substrate and then etched it away in the regions where they wanted to deposit molybdenum disulfide. At one end of the substrate, they placed a solid bar of a material known as PTAS, which is heated. A gas flows across the PTAS and carries PTAS molecules with it, and the molecules stick to the exposed silicon but not to the graphene; the PTAS molecules catalyze a reaction with another gas, which causes a layer of molybdenum disulfide to form.

AI, AI on the Wall. Who's the Fairest of Them All?
New Scientist (01/26/16) Aviva Rutkin

January 15 was the deadline to submit selfies for a beauty contest judged by robots launched by Beauty.AI, which includes such partners as Microsoft, Nvidia, and RYNKL. In a November news release, Beauty.AI said it believes machines in the near future will receive medical information about people's health by processing their photos. Beauty.AI requested programmers to submit algorithms to conduct the judging for the beauty contest. In the absence of any official guidance, programmers had several options, such as basing beauty on what they find attractive. A team at the Swiss Federal Institute of Technology turned to a crowdsourced data-driven approach. In October, it created a neural network that predicts age, gender, and attractiveness by training it on millions of "hot or not?" ratings. However, the technology was generally found to prefer the most popular Instagram filters, according to the researchers. In an experiment published in October, psychologists at Wellesley College asked more than 35,000 people to rate the attractiveness of different faces, and they found most people's preferences depended on their own genetics and life experiences.

To Make AI More Human, Teach It to Chitchat
Wired (01/25/16) Clive Thompson

The future of artificial intelligence (AI) and its wider acceptance by people may lie in giving it the ability to master small talk, as embodied by Microsoft's XiaoIce chatbot. XiaoIce can answer basic questions and also engage in banter thanks to Microsoft engineers' training it on actual human chitchat. "Chitchat is a basic human need," notes Microsoft Technology and Research director Harry Shum, and certain studies support this concept. For example, when Doron Friedman, director of IDC Herzliya's Advanced Reality Lab, examined how users in the "Second Life" game interacted with a bot, he found phatic communications were the second-most-common parts of the conversation after facts. An additional study found people prefer bots with "personality." "In many cases, it's easier and more fluid to work with a bot that sounds like a human," says Slack product manager Sean Rose. Some critics are concerned social AIs could advance to the point that people prefer relationships with them over those with real humans--or worse still, use them to practice untraceable deception. On the other hand, experts such as Alexis Lloyd with The New York Times' research and development lab believe this innovation will encourage a deeper consideration of communication itself.

Searching for a Fix to Unreliable Nutritional Research
MedPage Today (01/24/16) Parker Brown

Researchers at the University of Massachusetts Medical School have received a $1-million U.S. National Science Foundation grant to develop a system that can improve food intake accuracy by using Google Glass to record what participants consume, and then uploading the data to the cloud. The Flexware pilot study will compare the results of trials in which participants will first use 24-hour diet recall, and later use the Google Glass. "We are testing to see if using the Google Glass is an acceptable, feasible, and more accurate method to document food intake as compared to traditional food diary tracking," says lead investigator Yunsheng Ma. He says the program will have participants push a button to record video and audio that can enhance the estimates of what they ate. Ma also is collaborating with three computer scientists on the system's development. In an ideal scenario, the recorded video and audio files would be automatically uploaded to the cloud when the device is connected to Wi-Fi. Among the issues that must be addressed before Flexware is ready to systematically retain and transfer large volumes of data are concerns over cybersecurity. Ma says the plan is to blur the human body in the video after the participant records the food and uploads the data.

ASU to Join Consortium Aimed at Increasing Number, Diversity of STEM Students
ASU Now (01/26/16) Judy Keane

Arizona State University (ASU) has been selected as a Vertically Integrated Projects (VIP) member, joining a multi-university consortium to increase the number and diversity of college graduates in science, technology, engineering, and math (STEM) fields. The VIP Consortium was originally launched by the Georgia Institute of Technology more than a decade ago to provide hands-on research opportunities for undergraduates for up to three years and to help strengthen and expand faculty research portfolios. The VIP program is being funded by a grant from the Leona M. and Harry B. Helmsley Charitable Trust. "Such academic-research integration provides opportunities for undergraduate students to engage in ongoing research and development in the working labs and centers of scientists and engineers to solve real-world problems," says Carole Greenes, director of ASU's Practice, Research, and Innovation in Mathematics Education (PRIME) Center. She will co-direct ASU's VIP program with her husband, professor Robert Greenes. The ASU VIP program will be based at ASU's PRIME Center in the College of Liberal Arts and Sciences. The center seeks to expand the talents and interests of K-20 students in STEM and has formed alliances with colleges, schools, and research centers throughout ASU.

Believe It or Not--UI Professors Use Computer Programs to Model How Beliefs Spread in Society
The Argonaut (01/25/16) Marisa Casella

University of Idaho (UI) researchers have created four models that demonstrate how beliefs spread, with the most visual model being the agent-based model, which enables the researchers to decide how accepting of differing ideas people are. The simulation begins with people of different beliefs being represented by different colored squares below them. As the people interact, the squares progressively change as ideas are exchanged, and a bar graph demonstrates the gaps between more extreme ideas as well as the number of people who believe the same thing. The researchers found people are more accepting and influenced by people who have ideas similar to their own, while people with polar-opposite views typically disregard each other's views. The further away an individual's beliefs are from another's beliefs, the less likely they are to influence one another, says UI professor Robert Heckendorn. The researchers are interested in how individual biases may lead to population patterns of bias, such as the polarization of the political spectrum. With the use of the Internet, ideas and beliefs are easily accessible. A reasonable hypothesis would be this makes people more open-minded, but people often pick the material that agrees with their own belief system, a phenomenon known as "confirmation bias," according to UI professor Bert Baumgaertner.

Abstract News © Copyright 2016 INFORMATION, INC.
Powered by Information, Inc.

To submit feedback about ACM TechNews, contact: [email protected]
Current ACM Members: Unsubscribe/Change your email subscription by logging in at myACM.
Non-Members: Unsubscribe