Association for Computing Machinery
Welcome to the October 6, 2014 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 Web Privacy System Could Revolutionize the Safety of Surfing
University College London (10/06/14) Bex Caygill

Researchers at University College London (UCL), Google, Stanford University, Chalmers, and Mozilla Research say they have developed an open source system that protects Internet users' privacy while increasing the flexibility for Web developers to create applications that combine data from different websites, thereby improving the safety of surfing the Web. The system, called Confinement with Origin Web Labels (COWL), works with Mozilla's Firefox and the open source version of Google's Chrome Web browsers and prevents malicious code in a website from leaking sensitive information to unauthorized parties. "COWL achieves both privacy for the user and flexibility for the Web application developer," says UCL professor Brad Karp. "Achieving both these aims, which are often in opposition in many system designs, is one of the central challenges in computer systems security research." COWL helps block the development of Web applications that synthesize content from multiple websites, a technique that previously involved having the Same Origin Policy force Web developers to make design choices that put users privacy at risk. "What we've achieved in COWL is a system that lets Web developers build feature-rich applications that combine data from different websites without requiring that users share their login details directly with third-party Web applications," says Stanford Ph.D. student Deian Stefan.

Technology Takes the Wheel
The New York Times (10/05/14) Aaron M. Kessler

Although the fully autonomous vehicle is still the stuff of science fiction, autonomous and semi-autonomous driving functions are very much a reality and will be making their way into new vehicles in the next few years. Scott Belcher, president of the Intelligent Transportation Society of America, says 2014 has been a turning point, with major automakers buying into the concept of semi-autonomous cars in a big way. Analysts say most semi-automated driving initially will be confined to the highway, and several major manufacturers plan to have features in their cars that allow for autonomous highway driving within the next five years. Honda recently demonstrated its advances in automated driving technology on the highways around Detroit. The car autonomously steered through curves, merged into traffic, and took an exit off the highway. Meanwhile, French company Valeo, which makes assisted parking technology, is developing technology that would act like a robotic valet, enabling a car to find a spot and park itself after dropping off its passengers. Other automated features car makers are looking to add to their vehicles include safety features that would monitor a driver for signs of drowsiness or seize control and apply the brakes or steer out of the way if a crash is imminent.
View Full Article - May Require Free Registration | Return to Headlines | Share Facebook  LinkedIn  Twitter 

Carnegie Mellon's Mary Shaw Will Receive National Medal of Technology and Innovation
Carnegie Mellon News (PA) (10/03/14) Byron Spice

Carnegie Mellon University (CMU) computer science professor Mary Shaw is one of eight winners of the National Medal of Technology and Innovation announced by President Obama on Oct. 3 along with this year's 10 winners of the National Medal of Science. Shaw has taught computer science at CMU since 1971, where she took a leading role in the development of the field of software architecture. Her work has helped establish software architecture as a critical element of developing software for everything from automotive systems to international banking. Shaw also was one of CMU's earliest computer science Ph.D. graduates and has been a leader in developing the school's computer science curriculum, including the development of a graduate program targeted specifically at software professionals. "Building the reliable software systems that are the bedrock of commerce and communication today would not be possible without the engineering principles for large-scale software architecture pioneered by Mary and her colleagues at Carnegie Mellon," says CMU president Subra Suresh. Shaw has served as chief scientist of the university's Software Engineering Institute and as associate dean for professional education, and was a co-recipient of the ACM Special Interest Group on Software Engineering (SIGSOFT) Outstanding Research Award in 2011.

Advances in Computer Mobility, Connectivity, and Networks
National Science Foundation (10/02/14) Marlene Cimons

Williams College researchers are developing programs and platforms that will ensure systems work despite power outages and other issues associated with mobility and connectivity. "If you are writing software and you want to run it on machines not physically located in one place, you have to go through many steps, including plugging in your code, installing your software, configuring the machines," notes Williams College professor Jeannie Albrecht. She wants to develop ways to deal with anticipated connectivity failures, both when they are predictable and when they are not, and to examine methods for coping with delays they may produce. "We are looking at techniques for coping with the temporary unavailability of resources," Albrecht says. The researchers also are studying how to incorporate advance knowledge of connectivity failures into the program. "We've developed a couple of algorithms that make it easier to predict when we should abort," Albrecht says. She notes the research has the potential to greatly improve how smartphones, tablets, and other mobile devices function. The U.S. National Science Foundation is underwriting Albrecht's work with a five-year, $400,000 grant.

Google Working on Large-Scale Video Displays
The Wall Street Journal (10/03/14) Rolfe Winkler; Alistair Barr

Google's advanced-projects lab, Google X, is developing a giant display consisting of smaller screens that fit together to create a seamless image. The modular pieces can be arranged to form different sizes and shapes. The large screen could be used to watch TV or movies, browse the Internet, and read email, perhaps simultaneously, says NPD DisplaySearch research director Riddhi Patel. However, the project is still in the early stages of development because of the technical challenges associated with building such large screens and making the borders between the screen modules appear seamless. "The big challenge is to electronically, and through software, do the stitching between the seams," according to a person familiar with the project. Google X currently is trying to recruit more display experts to work on the problem. The project is being led by Mary Lou Jepsen, a former Massachusetts Institute of Technology professor who co-founded the One Laptop Per Child project. Microsoft also reportedly is working on big-screen display technology, in a project led by Perceptive Pixel founder Jeff Han. In 2012, Han predicted giant displays would eventually be in every meeting room, conference room, and classroom.
View Full Article - May Require Paid Subscription | Return to Headlines | Share Facebook  LinkedIn  Twitter 

Researchers Create Software for Google Glass That Provides Captions for Hard-of-Hearing Users
Georgia Tech News Center (10/02/14) Jason Maderer

Researchers at the Georgia Institute of Technology (Georgia Tech) have developed speech-to-text software for Google Glass to help people with hearing loss conduct conversations more easily. Users wear Google Glass while a second person speaks directly into a smartphone with an Android phone app. The speech is converted to text, sent to Glass, and shown on its display. The Captioning on Glass technology lets users "focus on the speaker's lips and facial gestures," says Georgia Tech professor Jim Foley. "I can glance at the transcription, get the word or two I need, and get back into the conversation." The project is being led by Thad Starner, head of the Contextual Computing Group. Starner says the phone-to-Glass system is helpful because speakers are more likely to construct their sentences more clearly. The smartphone software also enables the speaker to edit mistakes and send the changes to the person wearing the device. Foley and his team also are working on technology to enable two-way translations in English, Spanish, French, Russian, Korean, and Japanese.

Relationship Mining on Twitter Shows How Being Dumped Hurts More Than Dumping
Technology Review (10/01/14)

Venkata Garimella and colleagues at Aalto University in Finland are using data mining to examine the nature of relationship break-ups on Twitter. The key discovery was a process the team call "batch unfriending." Garimella and her team say there is clear evidence that after a break-up, the partners' number of friends and followers decline by about 15-20. They say this sudden change in each partners' network of connections and other signals could potentially lead to the creation of "early breakup warning" systems. The researchers used a dataset of 80 percent of all the tweets posted on Twitter during a 28-hour period in July 2013. They identified users who had mentioned another user in their profile along with certain words and phrases, including boyfriend, girlfriend, love, bf, and gf. They further filtered for English language users and excluded married couples and people in same-sex relationships, because psychological evidence shows both groups follow different dynamics. A total of 661 couples appeared to have broken up during the study period, and an additional 661 couples were randomly selected as a control. The researchers created word clouds of before and after word frequencies to highlight differences. Garimella and her team were surprised by the public nature of the conflicts that occurred following the break-ups.

Cybertools Offer New Channels for Free Speech, but Grassroots Organizing Still Critical
UC Berkeley NewsCenter (10/02/14) Sarah Yang

Although social media is driving free speech expression, the human factor in the form of grassroots organization is still vital. "Social media, purely as a logistical organizing mechanism, has transformed the ability of people to congregate and protest because of the speed and ease with which information is shared," notes the Center for Information Technology Research in the Interest of Society's Camille Crittenden. "Compared with the 1960s, you're going to be able to organize more quickly, and you're going to be able to document the event more comprehensively." However, Crittenden stresses the importance of physical presence and commitment, as does author Evgeny Morozov. "Digital tools are, and social change continues to involve many painstaking, longer-term efforts to engage with political institutions and reform movements," Morozov says. Meanwhile, the Berkeley Center for Law and Technology's Deirdre Mulligan warns technological savvy is not restricted to free-speech advocates; despots and oppressive regimes also can use the Internet and other cybertools to monitor and crack down on dissidents. To address such issues, researchers at the University of California, Berkeley and De Novo Group are developing technology to help citizens of oppressive governments bypass regime-imposed communication shutdowns. De Novo co-founder Yahel Ben-David says preserving anonymity is critical to setting up communication networks amidst such outages.

Online Crowd Can Guess What You Want to Watch or Buy
New Scientist (10/02/14) Hal Hodson

Crowds of online workers can be used to guess human preferences where there is little data to work with, according to the University of Illinois at Urbana-Champaign's Peter Organisciak and colleagues. The researchers hired crowdsourcing workers on Mechanical Turk and had them make personal recommendations. They presented 100 different salt and pepper shakers and 100 photos of different types of meals to the crowdsourcing workers and asked them to give each one a suitability rating out of five for a target person. The team only provided the workers with a small sample of the individual's actual taste in shakers and food. The researchers report the crowdsourcing workers did well in that the average rating from the top three recommenders matched the target person's own ratings to within half a star. Organisciak says websites such as Amazon and Netflix use algorithms to guess what consumers want, but they need to learn from large amounts of data. He notes using crowds instead of algorithms to determine preference is useful in personal data sets for which training an algorithm is impossible. The findings will be presented at the Conference on Human Computation & Crowdsourcing in Pittsburgh in November.
View Full Article - May Require Free Registration | Return to Headlines | Share Facebook  LinkedIn  Twitter 

Computer Science Times Are A-Changing
Cornell Chronicle (10/02/14) Bill Steele

Cornell University is celebrating the 50th anniversary of its Department of Computer Science, which recently held a two-day symposium that featured Juris Hartmanis, Edmund M. Clarke, and John Hopcroft, three computer science pioneers who are ACM A.M. Turing Award laureates. Carnegie Mellon University professor and Cornell graduate Edmund M. Clarke spoke about the lure of a school that had one of the first computer science departments. Cornell professor Juris Hartmanis noted there was not much interest in what they were doing then. Fellow Cornell professor John Hopcroft said the initial goal was "making computers useful," so there was a concentration on programming, but Cornell quickly shifted to theoretical computer science. Numerous computer science faculty members have joint appointments in other fields, and about 50 percent of all students graduating from Cornell have taken at least one computer science course. Hopcroft sees computer science undergoing a fundamental change, perhaps due to the ubiquity of computing technology. "Our department may have a leadership role in these changes," he said.

Computer Science and Statistics Team Awarded $1 Million to Make Big Data More User-Friendly
Virginia Tech News (09/26/14)

Virginia Polytechnic Institute and State University (Virginia Tech) researchers are working to make big data analytics usable and accessible to non-specialist, professional, and student users, thanks in part to a $1-million award from the U.S. National Science Foundation. The team is combining human-computer interaction with complex statistical methods to create a scalable and interactive system. The platform, known as Andromeda, makes use of a spatial metaphor that places similar objects, such as text documents and multidimensional data vectors, in closer proximity. When users recognize some objects in the space, the system is able to learn which data features express relevant patterns of similarity. Users gain insights by observing the updated structure of the visualization and by learning which features are most responsible for their injected feedback. The researchers say the platform essentially hides all the cryptic "knobs and controls" and provides a meaningful picture that users can manipulate. "We will enable people to interact with the data in order to identify novel 'what if?' questions," says Virginia Tech professor Chris North. "People can apply their knowledge about the subject area and recognize interesting, new patterns when they arise. The visualization is then tailored to how a person thinks about the data."

Google Lets Connected Devices Commune Without Specialized Apps
Network World (10/02/14) Jon Gold

A new initiative by Google seeks to form an open standard for the Internet of Things, dubbed the Physical Web by the company's Chrome team. The concept involves a subscribed discovery service that allows smart devices to distribute URLs to other smart devices in a given area. Users could potentially interact with vending machines, posters, and bus stops using Bluetooth and Web technology in the absence of any apps. "The number of smart devices is going to explode in number, both in our homes and in public space," predicts Google's Scott Jenson. "But the overhead of installing an app for each one just doesn't scale. We need a system that lets someone walk up and use a device with just a tap." To curb concerns about privacy, users' devices are never directly linked, so their locations are never revealed. Google also emphasizes the absence of active notifications. "The user will only see a list of nearby devices when they ask," Jenson says. Google pledged to treat the Physical Web project as an open Web standard.

Abstract News © Copyright 2014 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