Study Shows How Spammers Cash In
BBC News (11/10/08)
Researchers at the University of California, Berkeley and the University of California, San Diego (UCSD) hijacked a working spam network to analyze its economic value. The analysis found that spammers can make money by getting just one response for every 12.5 million emails sent. However, the researchers say that spam networks may be susceptible to attacks that make it more costly to send junk email. The researchers, led by UCSD professor Stefan Savage, took over a piece of the Storm spam network and created several proxy bots that acted as conduits of information between the command and control system for Storm and the hijacked home PCs that send the junk mail. The researchers used the machines to conduct their own fake spam campaigns. Two types of fake spam were sent, one that mimicked the way Storm spreads using viruses and the other aimed at tempting people to visit a fake pharmacy site and buy an herbal remedy to boost their libido. The fake pharmacy site always returned an error message when potential buyers clicked a button to submit their credit card details. The researchers sent about 469 million spam messages, and after 26 days and almost 350 million email messages, only 28 "sales" were made. The response rate for the campaign was less than 0.00001 percent, and would have resulted in revenues of $2,731.88, just over $100 a day for the measurement period. The researchers say that spam's small profit margin indicates that spammers would be economically susceptible to any disruptions in their networks.
Monday, November 10, 2008
Blog: Study Shows How Spammers Cash In
Saturday, November 8, 2008
Blog: Computers checking mathematical proofs?
November 8th, 2008
Computers checking mathematical proofs?
Computer-assisted of mathematical proofs are not new. For example, computers were used to confirm the so-called ‘four color theorem.’ In a short release, ‘Proof by computer,’ the American Mathematical Society (AMS) reports it has published a special issue of one its journals dedicated to computer-aided proofs. ‘The four Notices articles explore the current state of the art of formal proof and provide practical guidance for using computer proof assistants.’ And as said one of the researchers involved, ’such a collection of proofs would be akin to the sequencing of the mathematical genome.’ But read more…
Let’s first look at how mathematical theorems were proven in the past. “When mathematicians prove theorems in the traditional way, they present the argument in narrative form. They assume previous results, they gloss over details they think other experts will understand, they take shortcuts to make the presentation less tedious, they appeal to intuition, etc. The correctness of the arguments is determined by the scrutiny of other mathematicians, in informal discussions, in lectures, or in journals. It is sobering to realize that the means by which mathematical results are verified is essentially a social process and is thus fallible.”
This is why the concept of ‘formal proof’ is now being used. “To get around these problems, computer scientists and mathematicians began to develop the field of formal proof. A formal proof is one in which every logical inference has been checked all the way back to the fundamental axioms of mathematics. Mathematicians do not usually write formal proofs because such proofs are so long and cumbersome that it would be impossible to have them checked by human mathematicians. But now one can get ‘computer proof assistants’ to do the checking. In recent years, computer proof assistants have become powerful enough to handle difficult proofs.”
Here is a link to this special issue of the Notices of the American Mathematical Society (December 2008, Volume 55, Issue 11).
Thursday, November 6, 2008
Blog: Proof by Computer
Proof by Computer
EurekAlert (11/06/08)
New developments in the use of formal proof in mathematics are investigated by a series of articles by leading experts, published in the December 2008 edition of the Notices of the American Mathematical Society. The traditional method of proving theorems involves mathematicians presenting the argument in narrative form, and the correctness of the arguments is decided by the analysis of other mathematicians, in informal discussions, in lectures, or in journals. Problems of reliability inevitably crop up because the process of verifying mathematical results is basically social and therefore prone to error, and so the field of formal proof was developed to circumvent these shortcomings. A formal proof is one in which every single logical presumption has been checked all the way back to the basic mathematical precepts, and technology has advanced in recent years so that computer proof assistants that perform such checking are sufficiently powerful to accommodate challenging proofs. If the use of these assistants proliferates, the practice of mathematics could be dramatically revised. One vision is to have formal proofs of all central mathematical proofs, which Thomas Hales of the University of Pittsburgh compares to "the sequencing of the mathematical genome." The quartet of articles analyzes the current state of the art of formal proof and offers practical guidance for employing computer proof assistants.
Wednesday, November 5, 2008
Blog: Yahoo's Hadoop Software Transforming the Way Data Is Analyzed
Yahoo's Hadoop Software Transforming the Way Data Is Analyzed
SiliconValley.com (11/05/08) Ackerman, Elise
Yahoo!'s Hadoop open source data-mining program is capable of searching through the entire Library of Congress in less than 30 seconds. Universities also are using Hadoop, which is part of Yahoo!'s huge computing grid. "It makes it possible to actually take advantage of all the computers we have hooked up together," says Yahoo!'s Larry Heck. Hadoop improves the relevance of ads Yahoo! displays on the Internet by analyzing the company's endless flow of data, which is now more than 10 terabytes a day, in real time. As users navigate through Yahoo!, Hadoop determines which ads are likely to catch their attention. Yahoo! also will be using Hadoop on the sites owned by the 796 members of a newspaper consortium that is working with Yahoo! to sell more advertising at better prices. Hadoop was first used to build Yahoo!'s Web index. Since then, the software has been adjusted by engineers and researchers both inside and outside of the company for use in experiments with giant data sets. Amazon, Facebook, and Intel developers are using Hadoop for tasks such as log analysis to modeling earthquakes. "We are leveraging not only the contribution that we are giving to the software, but the contributions from the larger community as well, everybody wins from it," Heck says.
Tuesday, November 4, 2008
Blog: Multicore: New Chips Mean New Challenges for Developers
Multicore: New Chips Mean New Challenges for Developers
IDG News Service (11/04/08) Krill, Paul
The development of multicore processors is forcing software developers to work on getting software to be processed across multiple cores to fully utilize the new performance capabilities available in the hardware. However, this task has proven to be quite challenging, with developers struggling with issues surrounding concurrency and potential performance bottlenecks. Already, 71 percent of organizations are developing multithreaded applications for multicore hardware, according to a recent IDC survey. IDC analyst Melinda Ballou says developers need to approach multicore with a level of commitment to better practices throughout an organization and from a project perspective. Multicore processors are becoming increasingly common as single-core chips reach their limits and as power-consumption issues become more important. As hardware continues to change, the pressure will be on software developers to adapt and capitalize on the new capabilities. Developers must learn new techniques and use new tools to maximize performance. Intel's James Reinders acknowledges that developing multicore applications requires a much more complicated thought process about software design than most developers understand. "By and large, the majority of programmers don't have experience with it and are in need of tools and training and so forth to help take advantage of it," Reinders says. Intel is offering its Threading Building Blocks template library to help C++ programmers with parallel programming. The Intel Thread Checker helps find nondeterministic programming errors, and the Intel Thread Profiler helps visualize a program to check what each core is doing.
Monday, November 3, 2008
Blog: Microsoft Security Intelligence Report for First Half of 2008 (November 3, 2008)
SANS NewsBites Vol. 10 Num. 87 (fwd)
Tue, 4 Nov 2008
--Microsoft Security Intelligence Report for First Half of 2008 (November 3, 2008)
According to Microsoft's most recent semi-annual Security Intelligence Report, while machines running Windows Vista are less likely to be infected with malware than their Windows XP counterparts, ActiveX browser plug-ins still pose a threat to the newer operating system.
During the first six months of 2008, for each thousand times Microsoft's Malicious Software Removal Tool (MSRT) was executed, it scrubbed malware from three Vista SP1 machines, 10 Windows XP SP2 machines and eight Windows XP SP3 machines. Of the top 10 browser-based attacks against Vista during that same period, eight were ActiveX vulnerabilities. The report also found that 90 percent of disclosed vulnerabilities were in applications, while just six percent were in operating systems.
http://news.cnet.com/8301-1009_3-10080428-83.html?part=rss&subj=news&tag=2547-1009_3-0-20
[Editor's Note (Pescatore): There are far more applications than there are operating systems, so that last bit is not very surprising. The most meaningful data in this report is the chart that shows what types of installed malware the MSRT found and removed. It shows that Trojans and "potentially unwanted software" are getting through desktop defenses pretty easily - the signature and patch-centric approach to protecting desktops isn't dealing with the new, targeted threats that aim at the user, not unpatched PCs.]
Blog: Profile: Luis von Ahn; human-assisted computation
Profile: Luis von Ahn
BusinessWeek (11/03/08) Scanlon, Jessie
Carnegie Mellon University computer science professor Luis von Ahn has developed digitization software that could put the New York Times' entire archive, which dates back to 1851, online by late 2009. The newspaper has been using typists to digitize its archive, and in 10 years they have been able to digitize 27 years of articles. Von Ahn's software will process 129 years in less than 24 months. Von Ahn's research focuses on what he calls "human computation." He develops Web-based programs that take advantage of human abilities, such as reading or knowing common-sense facts, and then aggregating that knowledge to solve large-scale, ongoing problems in computer science. Von Ahn's first breakthrough technology, the Captcha, developed with von Ahn's thesis advisor Manuel Blum in 2000, is used by 60,000 Web sites to verify that the entity filling out a Web registration form is in fact a human. However, von Ahn released ReCaptcha, an updated version of the technology that replaced Captcha's random letters with words from library archives, and now the New York Times' archive, that computers were not able to read, helping complete digitization projects. Another example of human computation is von Ahn's ESP Game, in which two players are shown the same image and asked to type in descriptive labels. When the labels match, the players are awarded points and shown another image. The game helps generate tags for online images.