Association for Computing Machinery
Welcome to the January 18, 2013 edition of ACM TechNews, providing timely information for IT professionals three times a week.

Please note: In observance of the Martin Luther King, Jr. Day holiday, TechNews will not be published on Monday, Jan. 21. Publication will resume on Wednesday, Jan. 23.

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


Scientists Develop Device for Image Compression
New York Times (01/17/13) John Markoff

Duke University researchers are using a new class of artificial materials, called metamaterials, to design a sensor that compresses images much more efficiently than existing technologies. Metamaterials have qualities that bend light, X-rays, and radio waves in unusual ways. The researchers say the scanning sensor captures both still and video images while simplifying compression by integrating it directly into the sensor array. The technology permits image compression to be performed directly by the sensor hardware, instead of by the specialized hardware and software in use today. Although metamaterials might offer high compression ratios, the real advantage is in the potential for reductions in size, says Duke's John Hunt. For example, the most advanced planes and boats use a mechanically steered dish antenna for radar, which requires setting aside a large space to swivel the dish. "Our system could potentially replace that with a flat sheet wrapped onto the side of the fuselage," Hunt says. Depending on the wavelength the metamaterials are focused on, they are made with either printed circuitboards or semiconductors. The sensor elements can be laid out in a linear array or as a three-dimensional matrix.

In Wake of Earthquake, Researchers Envision a Cyber-Physical Cloud
Government Computer News (01/16/13) Rutrell Yasin

The U.S. National Institute of Standards and Technology (NIST) and Japan's National Institute of Information and Communications Technology (NICT) are collaborating to develop an eco-social system based on the interaction of billions of networked devices and people. The two organizations recently signed a memorandum of understanding to promote collaborative projects such as the development of a cyber-physical data cloud that collects, archives, organizes, manipulates, and shares very large cyber and physical data sets. After the massive earthquake that struck Japan in 2011, the Japanese government realized that cloud computing "is not only for cyberspace because it also needs to interact with the physical infrastructure and systems," says NICT researcher Kyoung-Sook Kim. Advances in broadband networks, mobile computing, and social media are laying the foundation for cyber-physical data clouds. NICT also is developing the cyber-physical system sensor information system, which will provide access to and analysis of sensor data and produce actionable information via visualization tools. NIST recently hosted a Cloud and Big Data workshop, which focused on how the two technologies can be used together, and big data's relation to and influence on cloud platforms and computing.

Facebook Shatters the Computer Server Into Tiny Pieces
Wired News (01/16/13) Cade Metz

Facebook hardware guru Frank Frankovsky is developing a blueprint for a new type of computer server that enables the user to add or remove the processor. "By modularizing the design, you can rip and place the bits that need to be upgraded, but you can leave the stuff that’s still good," Frankovsky says. The server design is part of the Open Compute Project, which aims to significantly reduce the cost and the hassle of maintaining and upgrading computer servers. "[The Open Compute Project] is about empowering the user to take control of infrastructure design," Frankovsky says. The modular processor specification is an extension of earlier hardware design open sourced by Facebook. The common slot used by Intel and AMD processor sockets is based on the PCIe connector used in existing servers. "We want to better match how the software is going to exercise the hardware," Frankovsky says. Facebook also has open sourced two other server designs. One is the latest version of the Facebook Web server, and the other is the company's first custom-built database server. Both systems are designed to reduce costs by stripping the hardware to the bare essentials, not using a hard drive, and running entirely on flash memory.

Web Hunt for DNA Sequences Leaves Privacy Compromised
New York Times (01/18/13) Gina Kolata

Online genetic data is raising privacy concerns after Whitehead Institute human genetics researcher Yaniv Erlich used the Internet to quickly pinpoint not only specific individuals who submitted DNA to the 1000 Genomes Project but also their family members who are not participating in the study. Using only strings of DNA, an age, and the participant's state, all of which were available through the project, Erlich was able to match participants to relatives who submitted DNA to public genealogy databases that offered surnames, and then to uncover entire family trees using Google. Study participants agreed to terms that did not guarantee their privacy, but Baylor College of Medicine lawyer Amy L. McGuire says no one anticipated that individuals would be identified. “To have the illusion you can fully protect privacy or make data anonymous is no longer a sustainable position," McGuire says. A strong supporter of data sharing, Erlich did not release the names that he discovered, but published his study in the journal Science. Privacy concerns should not be exaggerated because genetic information has been widely available online without instance of illicit identification, says the University of Utah's Jeffrey R. Botkin, although he is troubled by the publication of the identification methods.

Exploring the Baccalaureate Origin of Domestic Ph.D. Students in Computing Fields
Computing Research News (01/13) Susanne Hambrusch; Ran Libeskind-Hadas; Fen Zhao; et al.

The Computing Research Association Education Committee's Pipeline Project conducted a study of the baccalaureate origins of domestic U.S. students matriculating to computer science Ph.D. programs for the purpose of recruiting as well as improving the quality of the domestic pipeline. The research determined that about half of Ph.D. students come from 54 institutions of baccalaureate origin while the other half come from more than 747 institutions. Possibly the most significant finding of the research is that a small number of research universities and very selective colleges are the undergraduate institutions of origin for a large segment of domestic Ph.D. students and National Science Foundation Graduate Research Fellowship recipients. Yet many institutions send between one and five students to graduate school each year and their effect on the domestic pipeline is substantial. Master's institutions appear to be an underused source of prospective graduate students. Efforts to encourage graduate school applications from students in these schools benefit the pipeline. The research also found that many very successful schools are smaller and/or less well known, and disseminating some of these institutions' features could benefit the domestic Ph.D. pipeline.

Intel Preps Thin Fiber Optics to Shuffle Data Between Computers
IDG News Service (01/16/13) Agam Shah

Intel is developing thin fiber optics that will use lasers and light as a faster way to move data inside computers. The technology will be implemented at the motherboard and rack levels and use light to move data between storage, networking, and computing resources. The silicon photonics technology will be part of a new generation of servers that will need faster networking, storage, and processing subsystems, says Intel's Justin Rattner. He notes that silicon photonics could enable communication at speeds of 100 Gbps, and transfer data at high speeds while using less power compared to copper cables. The technology also could consolidate power supplies and fans in a data center, lowering component costs. After the infrastructure with silicon photonics is in place, server designs could change even further, Rattner says. Intel is working with Facebook to define new server technologies that will lead to the decoupling of computing, networking, and storage resources. Critical to this step is "the introduction of silicon photonics in not just the inter-rack fabric, but also the intra-rack fabric," Rattner says. "Over time you will see the server communication infrastructure, which includes switches, to include photonics," says analyst Dean McCarron.

Bad Grammar Make Good Password, Researchers Say
New Scientist (01/17/13) Hal Hodson

Computer users should not use good grammar for passwords, according to Ashwini Rao and colleagues at Carnegie Mellon University. The team has developed an algorithm that can crack long passwords that make grammatical sense as a whole phrase, even when interspersed with numbers and symbols. The algorithm is designed to make guesses by combining words and phrases from password-cracking databases into grammatically correct phrases. Ten percent of the long passwords that were tested were cracked exclusively using their grammar-sensitive methods. Other types of familiar structures such as postal addresses, email addresses, and URLs also are less secure passwords, even if they are long. The declining cost of processing power has made it more difficult for computer users to choose passwords that are easy to remember and secure. The team will present its research in February at the Conference on Data and Application Security and Privacy in San Antonio.
Share Facebook  LinkedIn  Twitter  | View Full Article - May Require Free Registration | Return to Headlines

Disparity Among First- and Second-Generation Immigrants in STEM Degrees
National Journal (01/15/13) Rosa Ramirez

Latino and Asian immigrants and their U.S.-born children are studying science, technology, engineering, and mathematics (STEM) subjects at different rates. About 25 percent of first- and second-generation Asian-Americans received degrees in STEM fields, compared to 14 percent of Latinos, according to a recent National Center for Education Statistics report. The two groups also differed from one another on socioeconomic characteristics, including age, low-income status, parents' education levels, and whether English was the primary language spoken at home, the study found. In addition, 25 percent of first- and second-generation Latinos major in humanities, social sciences, and general studies, compared to 21 percent of Asian-Americans. Finally, 46 percent of second-generation Asian-Americans took calculus in high school, compared to 21.5 percent of Latinos. Foreign students were excluded from the study, because they generally enter STEM fields at higher levels. The study concluded that those who complete advanced courses in high school are more likely to earn a bachelor’s degree. Meanwhile, a recent Census report shows that most degrees in computer, math, statistics, and engineering fields are still going to foreign-born residents, primarily to those from India and China.

Google Glass Hackathons Unveiled to Further Develop the Technology
eWeek (01/16/13) Todd R. Weiss

Google is planning two hackathon events to be held in San Francisco and New York City to showcase the Google Glass technology and gain input from developers who want to improve the system and make it more usable. "These hackathons are just for developers in the Explorer program, and we're calling them the Glass Foundry," Google says. "It's the first opportunity for a group of developers to get together and develop for Glass." The two-day events will provide a detailed look at the Google Mirror API, which is how Google Glass will interact with the user and exchange data. Google Glass consists of eyeglass-mounted computers with a side-mounted touchpad for controlling its various functions. The glasses will be able to display a wide range of views that can be catered to the user's interests. "For example, a person's name may be detected in speech during a wearer's conversation with a friend, and, if available, the contact information for this person may be displayed in the multimode input field," according to Google's patent application for the technology. Google Glass' basic elements include an Android-powered display, a small Webcam, a global positioning system locator, and an Internet connection node.

Mining Electronic Records for Revealing Health Data
New York Times (01/15/13) Peter Jaret

Electronic medical records could significantly advance medical science by enabling researchers to mine huge volumes of data at a low cost, circumventing clinical trials in some instances. "Medical discoveries have always been based on hunches," says Stanford University bioengineering and genetics professor Russ B. Altman. "This infrastructure makes it possible to follow up on those hunches." In 2011 Altman co-published a study linking the drug combination of antidepressant Paxil with the cholesterol drug Pravastatin to a spike in blood sugar. To find patients who fit his criteria, Altman used electronic medical records from Stanford University Medical Center, Harvard, and Vanderbilt. Obstacles to this type of research include the fact that all identifiable information must be removed from electronic records used for research due to patient privacy concerns. However, researchers must know whether they are looking at records for the same person, because a patient could be duplicated in multiple databases. Many academic medical centers and health organizations are removing identifiable data from records to make them available for research. "There's a growing sense in the field of informatics that we'll take lots of data in exchange for perfectly controlled data," Altman says.

IBM, Universities Team Up to Build Data Scientists
InformationWeek (01/15/13) Jeff Bertolucci

IBM is increasing its efforts to support big data academic programs. IBM has forged new partnerships with Michigan State University, Northwestern University, Yale University, and the University of Southern California in the past 18 months. "We're working with universities to provide help with curricula, technology, and real-world projects to help them teach this technology, and to help students put it into practice," says IBM's Rich Rodts. IBM wants to help prepare students for careers in data analytics by tailoring their skills to the needs of today's business environment, creating a new type of data scientist. "When I look at a data scientist, I look at someone who has to have an understanding of why analytics is important--and more importantly, what data is needed to make that analysis relevant," Rodts says. Some of the real-world projects focus more on communication skills, which attempts to prepare students for presenting analytic output to CEOs. Last summer IBM launched an internship program that enabled students to work with its staff and clients on business projects that utilize t he Watson supercomputer's cognitive computing skills, including natural-language processing, machine learning, and hypothesis generation.

Building Electronics From the Ground Up
University of South Carolina (01/14/13) Steven Powell

University of South Carolina researchers are developing bottom-up methods for creating electronics. Electronics are traditionally created with a top-down approach, which uses a prefabricated template to establish the pattern. Although this method has been successful, it has become more challenging as electronics have shrunk. The researchers' bottom-up method involves working with the individual molecules that go onto a surface, and coaxing them to self-arrange into the desired patterns. The researchers, led by professor Chuanbing Tang, have fabricated nanoparticles of pure crystalline iron oxide with controlled size and spacing on silicon wafers by covalently incorporating a ferrocene moiety into a tri-block copolymer. If the different block sections are designed correctly, the blocks will self-aggregate when placed on a surface, and the aggregation can be tapped to generate desirable nanoscale patterns without requiring any templates. Embedding metals within nanoscale designs is essential for the fabrication of electronic devices, and Tang’s method is a step forward for the discipline. "The industry won’t replace top-down methods, but they plan to use bottom-up together with the existing top-down methods soon," Tang says.

Ray Kurzweil Lifts Lid on Google AI Project
ZDNet (01/11/13) Sam Shead

Google plans to launch an artificial intelligence (AI) project focused on helping computers better understand human language, says Ray Kurzweil, the company's recently appointed director of engineering. Kurzweil says the idea is to give computers the ability to understand the language they are reading, and Google would know at a semantically deep level what users are interested in. "It will know not just the topic ... [but] the specific questions and concerns you have," he says. The technology will take several years to develop and will be integrated into services such as Google Search, rather than act as a standalone product. "I envision some years from now that the majority of search queries will be answered without you actually asking," Kurzweil says. The technology would just know this is something that the user wants to see, he notes.

Abstract News © Copyright 2013 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