Welcome to the August 16, 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."
UNSW Brings Together Science and Art to Make Sense of Data
Computerworld Australia
Rohan Pearce
August 15, 2017


Researchers at the University of New South Wales (UNSW) in Australia are using a range of technologies at the new Expanded Perception and Visualization Interaction Center (EPICenter) to derive meaning from large datasets via a combination of science and aesthetics. UNSW's Luc Betbeder says the EPICenter's goal is to "align the art and design community with the more hard-edged scientific community, starting with medicine and expanding into engineering and science." Among the facility's installations is a cylindrical visualization environment where 56 screens offer a seamless and immersive high-resolution display in both two- and three-dimensional (3D) formats. The EPICenter also houses the DomeLab, a 360-degree, roof-mounted fulldome that functions as a screen for 3D projectors. The facility is exploring the use of multilayered data in visualizations. "Because it's a physical place, we can bring people there to physically interact with the data--the physicality of the EPICenter is a big part of it," Betbeder says.

Full Article

A group of toy robots This Robot Lab Has No Idea What Its Robots Are Doing
The Wall Street Journal
Melissa Korn
August 15, 2017


The Georgia Institute of Technology's (Georgia Tech) Institute for Robotics and Intelligent Machines supports the Robotarium, an arena where scientists can run experiments on remote-controlled machines. Georgia Tech's Magnus Egerstedt says unpredictability is a regular feature of the Robotarium, where swarms of ground and airborne robots are put through their paces. Researchers use these swarms to test search-and-rescue scenarios, model flight formations for the U.S. Department of Defense, and predict the interactions of fleets of autonomous cars. Cameras located throughout the arena record the trials so the researchers conducting them can see the results of their experiments. The Robotarium is scheduled to relocate to a much larger space, which will host experiments with 120 ground bots and eventually 60 aerial bots. Stanford University professor Mac Schwager notes robot malfunctions are not the machines' fault. "It's not as if they're trying to break free or they've developed a will of their own," Schwager says.

Full Article
*May Require Paid Subscription
The Maddeningly Simple Way Tech Companies Can Employ More Women
The New York Times
Katharine Zaleski
August 15, 2017


There is an ongoing debate about the reasons for the lack of diversity in the technology industry, including candidate pools that are mostly male, and superficial notions of what it means to be a "cultural fit" for the organization, writes Katharine Zaleski, president and a co-founder of PowerToFly.com. Zaleski says many companies are alienating qualified women who want to work for them, and who they would want to hire, especially during the interview process. Silicon Valley companies are enthusiastically funding science, technology, engineering, and math programs in schools and nonprofits are focused on diversity, with the goal of creating a broader pipeline of talent over the next decade. However, Zaleski notes these firms also are missing opportunities to make simple, immediate improvements by changing how they communicate with female job candidates. Citing as an example a tech executive who boasts his company supports nonprofit coding organizations that train women, and who says he���ll have a booth at the Anita Borg Institute���s Grace Hopper Celebration of Women in Computing, where many women first experience being in a tech space where women are the majority, Zaleski says many tech executives don���t give as much thought to some of the simplest determinants of how successful a company will be in hiring diverse candidates. She says changing the hiring process to include more women, having diverse interview panels, and permitting current female employees to speak to candidates about their experiences could help companies recruit and retain more women.

Full Article
*May Require Paid Subscription

Aftermath of a tsunami University Unveils Set of Tech Tools to Help Manage Disasters
The Nation (Thailand)
August 15, 2017


Researchers at Thammasat University (TU) in Thailand have developed two software tools they say could be useful in recovering from natural disasters, such as the recent floods in their country. The first tool is a website that can track and locate disaster victims in real time via their social media use. The site pulls information from a range of sources, and those with access to social media could seek aid by providing their location along with an agreed-upon hashtag describing the disaster. In addition, rescue teams could add information in real time to keep the website updated about the situation and the mission status. The second tool is a drone that can carry out three-dimensional inspections of disaster damage at ancient sites and other locations. Both tools have proven successful in tests, and TU professor Pakorn Sermsuk says they could be useful in disaster mitigation.

Full Article
Helping Computers to See
The Trinidad & Tobago Guardian
August 14, 2017


Naila Murray heads the Naver Labs Europe Computer Vision Group in France, where she develops software to help computers analyze and translate visual sensor input into usable, actionable information. "Computer vision development is all about replacing the brain as an interpretive system," Murray says, and her work seeks to span the gap between pixels and meaning. Murray says computer-vision research concentrates on human vision's reliance on learned knowledge so objects can be recognized in context, even when they cannot be seen clearly. She notes this represents a massive dataset of information that is intuitively collected. Murray says it currently is not very difficult to deceive computer-vision systems, as a simple overlay of image noise can completely disrupt a computer's view of an image and even lead it astray--a vulnerability with potentially deadly consequences. As for computer vision's ramifications for privacy and personal security, Murray hopes the technology will be regulated via legislation, public awareness, and good sense.

Full Article

Video player with ‘buffering’ message High-Quality Online Video With Less Rebuffering
MIT News
Adam Conner-Simons
August 14, 2017


Researchers at the Massachusetts Institute of Technology's (MIT) Computer Science and Artificial Intelligence Laboratory have developed Pensieve, an artificial intelligence (AI) system that uses machine learning to choose different algorithms depending on network conditions. The researchers say the algorithm offers a higher-quality streaming experience with less rebuffering than existing systems. During testing, the researchers found Pensieve could stream video with 10-percent to 30-percent less rebuffering than other approaches, and at levels that users rated 10-percent to 25-percent higher on key "quality of experience" metrics. In addition, Pensieve can be customized based on a content provider's priorities. The team tested Pensieve in several settings, including using Wi-Fi at a cafe and a network while walking down the street. "We're excited to see what systems like Pensieve can do for things like [virtual reality]," says MIT professor Mohammad Alizadeh. "This is really just the first step in seeing what we can do."

Full Article
'Organismic Learning' Mimics Some Aspects of Human Thought
Purdue University News
Emil Venere
August 14, 2017


Researchers at Purdue University and other institutions have developed "organismoids," a new computing technology that mimics certain aspects of human thought. Purdue professor Kaushik Roy says organismoids have the ability to "not only learn new information but...also learn what to forget." The project used a ceramic "quantum material" to create organismoid devices, which expand in reaction to hydrogen gas and contract when the gas is removed. Purdue professor Shriram Ramanathan says the expansion induces a reversible quantum mechanical effect, enabling electrical resistance to change by orders of magnitude. Ramanathan notes the degree of conduction and insulation can be precisely tuned, which resembles habituation. "The behavior of conductance going up and down in exponential fashion can be used to create a new computing model that will incrementally learn and at the same time forget things in a proper way," Roy says. He says the technology could lead to a new "neural learning model" called adaptive synaptic plasticity.

Full Article

Windshield of a school bus How Do You Fix a School-Bus Problem? Call MIT
The Wall Street Journal
Jo Craven McGinty
August 11, 2017


A team of researchers at the Massachusetts Institute of Technology (MIT) has developed an algorithm designed to plot more efficient Boston public-school bus routes, which won a competition sponsored by the Boston school system. The algorithm can devise a solution in about half an hour by optimizing multiple routes while also accounting for traffic, different-size buses, students with special needs such as wheelchair access, and staggered school days. The researchers initially paired student clusters with bus stops, and then, using Google Maps travel times to factor in traffic volume, they connected the bus stops into six to eight efficient route solutions for each school. "Next, we combined one solution from one school to another school and to another school to optimize the overall system," says MIT's Dimitris Bertsimas. Boston Public Schools says the algorithm could save as much as $5 million, 20,000 pounds of carbon emissions, and 1 million bus miles annually.

Full Article
*May Require Paid Subscription
De-Jargonizing Program Helps Decode Science Speak
American Technion Society
Kevin Hattori
August 10, 2017


Researchers at the American Technion Society and HIT-Holon Institute of Technology in Israel have developed the De-Jargonizer program to automatically identify terms the average person might not know. After a text is uploaded or pasted, the algorithm color-codes words in the text as either frequent or intermediate-level general vocabulary, or jargon. The code is based on the frequency of words on a public Internet news site. "The De-Jargonizer provides a grim glimpse at the current level of jargon in scientific writing," says Technion professor Ayelet Baram-Tsabari. The researchers found lay summaries include 10-percent jargon, compared to 14-percent jargon in academic abstracts on average. The program is designed to help researchers improve and adapt vocabulary use when communicating with non-experts. "The scientists intuitively understand they need to use less jargon when speaking with the public than to their peers, but using so many unfamiliar words excludes the very people they are trying to engage," Baram-Tsabari says.

Full Article
Teaching AI Systems to Behave Themselves
The New York Times
Cade Metz
August 13, 2017


A small but growing community of artificial intelligence (AI) researchers is beginning to explore mathematical techniques that aim to prevent machines from demonstrating unexpected, unwanted, and harmful behavior. Safety concerns are growing as AI moves into online services, security devices, and robotics. Researchers at OpenAI are developing algorithms that can learn tasks through hours of trial and error while receiving regular guidance from human teachers. The researchers believe algorithms that combine human and machine instruction can help keep automated systems safe. OpenAI is working to build reinforcement-learning algorithms that accept human guidance to ensure that systems do not stray from assigned tasks. Some researchers in the field are working to ensure that systems do not make mistakes on their own, while others aim to prevent hackers and other bad actors from exploiting AI flaws.

Full Article
*May Require Free Registration

Empty driver's seat of a car A Researcher Disguised Himself as a Car Seat to Teach Driverless Cars How to Communicate With Humans
The Washington Post
Michael Laris
August 9, 2017


Researchers at Virginia Polytechnic Institute and State University (Virginia Tech) are teaching autonomous vehicles to communicate with humans. The trials involve scientists costumed as car seats to give the appearance that no one is driving the vehicles while experiments are conducted. The researchers note the disguise renders drivers less visible while enabling them to safely monitor and respond to surroundings. The researchers say the study "is investigating the potential need for additional exterior signals on automated vehicles. This research is relevant for ensuring pedestrians, cyclists, and other drivers are accommodated." Scientists from Virginia Tech, Ford Motor Company, the University of Leeds in the U.K., and the National Institute of Advanced Industrial Science and Technology in Japan raised related issues at a July symposium. Such issues include how cars should communicate with others on the road while they are in motion, stopped, or transitioning, as well as whether standardizing such communication is sensible.

Full Article
*May Require Free Registration
First Evidence That Social Bots Play a Major Role in Spreading Fake News
Technology Review
August 7, 2017


Researchers at Indiana University in Bloomington say they have proof that social bots are the key to the online proliferation of fake news. The team tracked about 400,000 claims from websites that routinely publish fake news, and examined their spread through Twitter. In addition, they monitored about 15,000 stories from fact-checking groups and more than 1 million Twitter posts mentioning them. The Twitter accounts that spread this news were analyzed, and the researchers devised the Hoaxy platform to track fake news claims, and the Bolometer to determine whether the Twitter accounts were run by persons or bots. The team says bot-run accounts are more likely to actively spread misinformation, especially just after publication. Furthermore, such programs are designed to target tweets at influential users with highly connected nodes. "These results suggest that curbing social bots may be an effective strategy for mitigating the spread of online misinformation," the researchers say.

Full Article
*May Require Paid Subscription
I'm a Woman in Computer Science. Let Me Ladysplain the Google Memo to You.
Vox
Cynthia Lee
August 11, 2017


One of the main reasons the now-infamous Google "manifesto" has been met with so much support is its quasi-professional tone, writes Cynthia Lee, a lecturer in the Computer Science Department at Stanford University. Lee says many defenders are genuinely confused that a document that appears so dispassionate and reasonable could provoke such an emotional response. In addition, that response confirms the opinion in many male technology workers' minds that women are more emotional and less quantitative in their thinking. However, Lee says these arguments have been used before, and the manifesto was simply a tipping point for the backlash that followed. She says the technology industry has been plagued with endless skepticism that every woman faces, and many are exhausted with trying to constantly defend their presence. Lee also notes if the memo's proposals were adopted, some women wouldn't get to attend the Anita Borg Institute's Grace Hopper Celebration of Women in Computing, a conference that provides many women with their first experience of being in a majority-women tech conference space.

Full Article
SIGGRAPH Asia 2017 Conference
 
Subscribe to Communications of the ACM
 

Association for Computing Machinery

2 Penn Plaza, Suite 701
New York, NY 10121-0701
1-800-342-6626
(U.S./Canada)



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]