Association for Computing Machinery
Welcome to the June 11, 2012 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).


The AI Game That Knows You Better Than Anyone
New Scientist (06/06/12) Jacob Aron

Many computer science researchers are trying to create an artificial intelligence (AI) system that offers video game players a more tailored experience. For example, 22Cans developer Peter Molyneux plans to harness the massive amount of personal data available on social media to learn what players enjoy and create characters that connect with them as individuals. Molyneux will soon launch the first of 22 experiments designed to probe the psychology of social-media users, which will inform the final design for a game that will be released in the next two years. The first experiment, Curiosity, puts players in a virtual room containing a single black cube that they tap away at to open it with a secret predetermined sequence. Molyneux says opening the cube will reveal something "unique and amazing," and after its revealing 22Cans will study how news of the revelation spreads through social media. Meanwhile, Imperial College London researcher Michael Cook is using social media to teach games about the real world. Cook created an AI system that designs its own simple video games and recently added the ability to base these games on news articles using personal opinions gathered from Twitter users.

Penn Reaches Thousands Online with Coursera
Penn Current (06/07/12) Tanya Barrientos

Twelve University of Pennsylvania professors have agreed to be the first to teach courses as part of Coursera, a new online education platform. Coursera also is working with Princeton University, Stanford University, and the University of Michigan in its initiative to distribute not just math, engineering, and science courses to users, but also an assortment of courses in the humanities and social sciences, taught to thousands of students simultaneously. "We decided that we had to think about how to expand the scale of this, in terms of giving students a great education across disciplines," says Stanford professor Daphne Koller, who co-created Coursera with Andrew Ng. Koller and Ng received $16 million in venture capital from two major Silicon Valley investment companies in April, to go toward Coursera's rollout. Penn professor Robert Ghrist will teach a course in single variable calculus, which is expected to draw over 12,000 students from around the world. "I love the fact that I am going to be able to reach a lot of high school students who are perhaps bored with the treatment that calculus gets in high schools ... and professionals who are finding that they need to know this stuff for their jobs and may never have learned it properly the first time," he says.

Eleven Countries Use an Azti-Tecnalia Software as a Method to Assess the Status of Their Seabeds
Basque Research (06/05/12)

AZTI-Tecnalia has developed software that allows users to assess the ecological status of seabed fauna. AZTI's Marine Biotic Index (AMBI) includes data on more than 6,000 seabed animals worldwide. For an environmental study of benthos, the software would evaluate the status of fauna based on calculating the proportions of species sensitive to pollution, species indifferent to pollution, tolerant, and first- and second-order opportunistic species. The software classifies each of the species collected in samples taken from the seabed, based on these five environmental groups. The selection is particularly accurate, considering the volume of records on animal species, and the software can identify which of the ecological groups mentioned they belong to. In addition to offering AMBI, the research and development center has adapted a version specifically for the European Water Framework Directive. Eleven European countries are now using the software tool, which can be downloaded free of charge, and its use could spread to the United States, Chile, and China. The tools could help protect against marine and shoreline pollution, and they have aided publications, scientific reports, and Ph.D. theses in more 30 countries.

Assembling, Visualizing and Analyzing a Tree of All Life
National Science Foundation (06/04/12) Cheryl Dybas

The U.S. National Science Foundation's (NSF's) Assembling, Visualizing, and Analyzing the Tree of Life (AVAToL) program aims to build a comprehensive tree of life that brings together everything scientists know about how all species are related. The researchers are creating the infrastructure and computational tools to enable automatic updating of the tree of life, as well as developing the analytical and visualization tools to study it. Assembling the branches for all species of animals, plants, fungi and microbes will require new computational tools for analyzing large data sets, for combining diverse kinds of data, and for connecting vast numbers of published trees into a synthetic whole. AVAToL will enable researchers to go online and compare their trees to others that have already been published. The goal is to automatically incorporate new trees, so the complete tree can be continuously updated. The three NSF-funded AVAToL projects are Duke University's Automated and Community-Driven Synthesis of the Tree of Life, the University of Idaho's Arbor: Comparative Analysis Workflows for the Tree of Life, and SUNY-Stony Brook's Next Generation Phenomics for the Tree of Life.

IBM Partners With Syracuse University to Tap Next-Gen Mainframe Workers
eWeek (06/04/12) Darryl K. Taft

IBM announced a partnership with Syracuse University to help college students build smarter computing skills to manage both traditional and new systems in large global enterprises. Syracuse's Global Enterprise Technology (GET) curriculum is an interdisciplinary program focused on preparing students for successful careers in large-scale, technology-driven global operating environments. A consortium of technology partners, including IBM, supplies GET students with technology platforms and multiple systems experience. "These courses and the IBM technology platform help prepare students to build large global data centers, allow them to work across multiple systems, and ultimately gain employment in large global enterprises," says Syracuse professor David Dischiave. Almost 500 students have completed the GET minor since its inception. "This is part of our ongoing effort to build ecosystems around the mainframe," says IBM Rational's Charles Chu. "We're funding access to software, hardware and more for the University of Syracuse, and we're adding the mainframe to our toolset." In 2010, IBM introduced the zEnterprise system, which extends the strengths and capabilities of the mainframe to other systems and workloads running on AIX on Power7, Linux on System x, and Microsoft Windows. More than 120 new client worldwide have selected the IBM mainframe platform to serve as the backbone of their information technology infrastructure since the zEnterprise's unveiling, according to IBM.

Is a Rat Smarter than Google? That's What Two AI Experts Say
MSNBC (06/04/12) Sean Captain

Artificial intelligence (AI) researchers Yann LeCun and Josh Tenenbaum recently spoke at the World Science Festival about the relative capacity of modern AI technologies. In terms of computational ability, even the most-powerful computers in the world are just approaching that of an insect, according to LeCun. Even some of the seemingly amazing things that AI can do today, such as providing directions, use only a basic kind of intelligence called simple planning, Tenenbaum says. Real intelligence is using what has been learned to figure out new situations that have never been experienced. "That whole context of communication intelligence, of getting inside another person just by the data of what they say and you say back, that's the heart of human intelligence," Tenenbaum says. LeCun is experimenting with a driving robot that tries to identify objects around it. The researchers estimate that in order for a computer to match the human brain's computing capacity, it would have to complete 1 quintillion operations every second. The researchers estimate that supercomputers will reach this level within 100 years.

Slime Moulds Work on Computer Games
Inderscience Publishers (06/07/12)

Researchers at the University of the West of England are using slime to find ways to calculate the shape of a polygon linking points on a surface. The researchers, led by Andrew Adamatzky of the Unconventional Computing Center, show that computing a polygon defining a set of planar points is a classical problem in modern computing geometry. The slime mold Physarum polycephalum has a complicated life cycle with fruit bodies, spores, and single-cell amoebae, but in its vegetative, plasmodium, stage it is basically a single cell containing many nuclei. "Plasmodium's foraging behavior can be interpreted as computation, when data are represented by spatial configurations of attractants and repellents, and results by structures of protoplasmic network," Adamatzky says. The cultivation of plasmodium on a surface dotted with attractants and repellents should make it possible to grow a network of tubes that link the attractant points and avoid the repellents while generating the most effective connectivity between the former. The research into how slime molds compute suggest that this latest step could lead to a parallel embedded computing processor that uses non-linear chemical media instead of standard components such as silicon chips to carry out computations.

EastWest Institute Proposes Public Health Model for Internet Cybersecurity (06/06/12) Mickey McCarter

Cybersecurity professionals should establish a framework to view and monitor the Internet's health based on a global public health model, according to a recent EastWest Institute (EWI) report. "With this breakthrough report, we have the opportunity to treat the health of the entire Internet as a shared problem needing cooperative solutions," says EWI president John Mroz. Following a public health model, cybersecurity experts could track and block malware and stop malicious actors, according to the EWI report. "A public health agency, whether operating on the local, state, national, or international level, is a robust model for potential application to cyberspace, with basic functions including education, monitoring, epidemiology, immunization, and incident response," the report states. After establishing a cybersecurity health framework, Internet security researchers could apply the global health concepts of education, monitoring, epidemiology, immunization, and incident response to cybersecurity practices for the public good. "As use and reliance on the Internet continues to grow, improving Internet health requires all ecosystem members to take a global, collaborative approach to protecting people from potential dangers online," says Microsoft's Scott Charney.

Online Classes See Cheating Go High-Tech
The Chronicle of Higher Education (06/03/12) Jeffrey R. Young

As online classes grow in popularity, the issue of online cheating also may grow in prominence unless courses are designed carefully. Part of the solution involves combating cheating technology with better preventative technology. Blackboard, an education-software company, has developed learning-management software that features a service that checks papers for signs of plagiarism. Blackboard's John Fontaine is developing a new system that could establish a document fingerprint for each student when they turn in their first assignments, and notice if future papers differ in style in suspicious ways. Massachusetts Institute of Technology (MIT) researchers are looking for new ways to verify the identity of students online by analyzing each user's typing style to help verify identity. This type of electronic fingerprinting could be combined with face-recognition software to ensure accuracy, says MIT's Anant Agarwal. Recently, 100 researchers met at the University of Kansas at the Conference on Statistical Detection of Potential Test Fraud and agreed that groups that offer standardized tests, companies that develop anti-cheating software, and researchers need to work together and share their work. "It's important that the research community improve perhaps as quickly as the cheating community is improving," says University of Kansas professor Neal Kingston.

New Statistical Model Lets Patient's Past Forecast Future Ailments
University of Washington News and Information (06/04/12) Molly McElroy

University of Washington researchers have developed an algorithm that makes predictions based on what a patient has already experienced as well as the experiences of other patients showing a similar medical history, through analysis of medical records from thousands of patients. "This provides physicians with insights on what might be coming next for a patient, based on experiences of other patients," says Washington professor Tyler McCormick. What differentiates the model from others is that it shares information across patients who have similar health problems, which allows for better predictions when details of a patient's medical history are hard to come by, according to McCormick. The new algorithm can compare a patient's health problems with other patients who have a more extensive medical record that includes similar symptoms. "We're looking at each sequence of symptoms to try to predict the rest of the sequence for a different patient," McCormick says. The researchers used medical records taken from a multiyear clinical drug trial involving tens of thousands of patients aged 40 and older. "We hope that this model will provide a more patient-centered approach to medical care and to improve patient experiences," McCormick says.

Bionic Cells Can Do Basic Arithmetic
New Scientist (06/06/12) Jacob Aron

Martin Fussenegger at the Swiss Federal Institute of Technology has created biological versions of half adder and half subtractor digital circuits inside two sets of embryonic kidney cells. These are the most complex biological circuits ever created, and they could lead to more advanced computational circuits. Fussenegger's cells use erythromycin and phloretin as inputs, switching a reaction within the two types of cell on or off. The reaction leads to the production of a red or green fluorescent protein that signals the result of the calculation. These reactions take place without interfering with the cell's ordinary functions, allowing them to speak the binary language of computers while continuing to work as normal cells. The cell computers are more flexible than electronic systems because both the input molecules and the output proteins can be replaced with other biological signals, which means a computer could be designed to take a signal from an infection as its input, and the output would be to deliver an appropriate treatment. "The team have taken this to the next level by showing how one can encode decision-making logic into cells rather than just producing a response," says Manchester Metropolitan University's Martyn Amos.

Move Over Pie Charts, Here Come FatFonts
University of Calgary (05/30/12)

University of Calgary researchers have developed FatFonts, a new font style for numbers that represents their relative value. In FatFonts, the amount of ink used for each digit is proportional to its quantitative value, and FatFonts aims to align numeric value and visual representation. "The digit '3' uses three times the amount of ink than digit '1,'" says Calgary Ph.D. student Uta Hinrichs. "When you put FatFonts in a grid, they work like a table where you can read the numbers but also as an image that you can interpret visually." Researchers can use FatFonts for large numeric data sets where it is important to see both individual data values and an overview of the information. Former Calgary postdoc Miguel Nacenta, who conceived of FatFonts, notes "they're particularly interesting to use on large high-resolution wall displays or large posters. Stepping back provides you an overview image while walking closer to the display reveals the data values." Hinrichs says people have begun using FatFonts for their own visualizations and have even invented their own FatFont types.

Abstract News © Copyright 2012 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