- talks - lists.cs.princeton.edu

Today's Colloquium Cancelled
by Andrew W. Appel 18 Nov '09

18 Nov '09

Today's colloquium is cancelled. David Naumann cannot speak due to an illness.

1 0

David Naumann--Colloquium Speaker, Wednesday, October 18, 2009
by Michele J. Brown 16 Nov '09

16 Nov '09

Title: Smart Assertions for Dumb Provers David Naumann, Stevens Institute of Technology 4:30PM- Small Auditorium CS105 Host: Andrew Appel Assertions are widely used for runtime checking of software correctness, and increasingly used for compile time checking of simple properties like array index bounds. Structural integrity of pointer based data structures and design patterns often involves recursively defined properties which are costly to check at runtime and which appear to be unsuited to static checking using fully automated theorem provers. Remarkably, judicious use of ghost state lets programmers express recursive properties in pure first order logic and successfully use fast provers based on SAT solving. In this talk I will introduce the technique in elementary terms, for programmers, who may find it immediately useful in testing. I will also outline research questions about how to use the technique in frame conditions, in refactoring, in foundational proofs, in refinement types, in terms of aspects, and in checking information flow properties. Joint work with Anindya Banerjee, Mike Barnett, and Stan Rosenberg

1 0

Colloquium Speaker, Wednesday, November 18, 2009- David Naumann
by Michele J. Brown 13 Nov '09

13 Nov '09

Title: Smart Assertions for Dumb Provers David Naumann, Stevens Institute of Technology 4:30PM- Small Auditorium CS105 Host: Andrew Appel Assertions are widely used for runtime checking of software correctness, and increasingly used for compile time checking of simple properties like array index bounds. For pointer based data structures and design patterns, structural integrity conditions often involve recursively defined properties which are costly to check at runtime and which appear to be unsuited to static checking using fully automated theorem provers. Remarkably, judicious use of ghost state lets programmers express recursive properties in pure first order logic, thereby enabling the use of fast provers based on SAT solving. In this talk I will introduce the technique and pose research questions about how to use the technique in frame conditions, in foundational proofs, in refinement types for higher order programs, in terms of aspects, and in checking information flow properties. Joint work with Anindya Banerjee, Mike Barnett, and Stan Rosenberg

1 0

Fwd: [ens] talk about blackhole identification by Italo Cunha, Nov 11 11am, CS 302
by Yaping Zhu 09 Nov '09

09 Nov '09

Begin forwarded message: > From: Yaping Zhu <yapingz(a)CS.Princeton.EDU> > Date: October 7, 2009 6:55:55 PM PDT > To: ens(a)lists.cs.princeton.edu > Subject: [ens] talk about blackhole identification by Italo Cunha, > Nov 11 11am, CS 302 > > hi all, > Italo Cunha from UPMC Paris Universitas is visiting Princeton on Nov > 11. He will be giving a talk at 11am in CS 302 (title and abstract > below). please let me know if you'd like to meet with him in person. > > thanks, > Yaping > > Title: Measurement Methods for Fast and Accurate Blackhole > Identification with Binary Tomography > > Abstract: Binary tomography--the process of identifying faulty network > links through coordinated end-to-end probes--is a promising method for > detecting failures that the network does not automatically mask (e.g., > network "blackholes"). Because tomography is sensitive to the > quality of > the input, however, naive end-to-end measurements can introduce > inaccuracies. This paper develops two methods for generating inputs to > binary tomography algorithms that improve their inference speed and > accuracy. Failure confirmation is a per-path probing technique to > distinguish packet losses caused by congestion from persistent link or > node failures. Aggregation strategies combine path measurements from > unsynchronized monitors into a set of consistent observations. When > used > in conjunction with existing binary tomography algorithms, our methods > identify all failures that are longer than two measurement cycles, > while > inducing relatively few false alarms. In two wide-area networks, our > techniques decrease the number of alarms by as much as two orders of > magnitude. Compared to the state of the art in binary tomography, our > techniques increase the identification rate and avoid hundreds of > false > alarms. > > Bio: Ítalo Cunha received the B.Sc. degree in computer science and the > M.Sc. degree in computer science from Universidade Federal de Minas > Gerais, Brazil, in 2004 and 2007, respectively. He is currently a > second > year Ph.D. candidate in UPMC Paris Universitas, doing his research > with > Thomson, Paris, France. His research interests are in network > measurement, troubleshooting, and management. > _______________________________________________ > ens mailing list > ens(a)lists.cs.princeton.edu > https://lists.cs.princeton.edu/mailman/listinfo/ens

1 0

Colloquium Speaker, Wed. Nov. 11, 2009- Andrew Bavier
by Michele J. Brown 09 Nov '09

09 Nov '09

Title: Establishing Resource Allocation Policies in the PlanetLab Federation Andrew Bavier, Princeton University, Computer Science Department Wednesday, Nov 11, 2009 4:30PM Small Auditorium, CS 105 Over time PlanetLab has evolved from a centrally-administered testbed to a federation of regional testbeds that are operated independently. The PlanetLab software is also freely available, which has allowed many organizations to bring up their own "private" PlanetLabs. For most private PlanetLabs, a barrier to entering the federation is the inability to express access and allocation policies for their own resources, and the lack of tools to communicate, analyze, and enforce those policies. In the first part of my talk, I will present an overview of the PlanetLab federation. Next I will focus on VINI, a private PlanetLab we have deployed in Internet2 and NLR, and which is intended for evaluating new network architectures. Finally, I will describe a tool ("sfatables") we have built for expressing and enforcing resource allocation policies. Our approach was designed to enable VINI to federate with PlanetLab, but it is general enough to accommodate a diverse collection of computing facilities. This talk describes joint work with Larry Peterson, Sapan Bhatia, Jennifer Rexford, and Nick Feamster.

1 0

Colloquium Speaker, Wed. Nov. 11, 2009- Andrew Bavier
by Michele J. Brown 05 Nov '09

05 Nov '09

Title: Establishing Resource Allocation Policies in the PlanetLab Federation Andrew Bavier, Princeton University, Computer Science Department Wednesday, Nov 11, 2009 4:30PM Small Auditorium, CS 105 Over time PlanetLab has evolved from a centrally-administered testbed to a federation of regional testbeds that are operated independently. The PlanetLab software is also freely available, which has allowed many organizations to bring up their own "private" PlanetLabs. For most private PlanetLabs, a barrier to entering the federation is the inability to express access and allocation policies for their own resources, and the lack of tools to communicate, analyze, and enforce those policies. In the first part of my talk, I will present an overview of the PlanetLab federation. Next I will focus on VINI, a private PlanetLab we have deployed in Internet2 and NLR, and which is intended for evaluating new network architectures. Finally, I will describe a tool ("sfatables") we have built for expressing and enforcing resource allocation policies. Our approach was designed to enable VINI to federate with PlanetLab, but it is general enough to accommodate a diverse collection of computing facilities. This talk describes joint work with Larry Peterson, Sapan Bhatia, Jennifer Rexford, and Nick Feamster.

1 0

Colloquium Speaker- tomorrow-10/29
by Michele J. Brown 28 Oct '09

28 Oct '09

Title: Network-Level Spam and Scam Defenses Professor Nick Feamster, Georgia Tech Host: Jennifer Rexford CS 105-Small Auditorium 4:30pm, Thursday This talk introduces a new class of methods called "behavioral blacklisting", which identify spammers based on their network-level behavior. Rather than attempting to blacklist individual spam messages based on what the message contains, behavioral blacklisting classifies a message based on how the message itself was sent (spatial and temporal traffic patterns of the email traffic itself). Behavioral blacklisting tracks the sending behavior of an email sender from a wide variety of vantage points and establishes "fingerprints" that are indicative of spamming behavior. Behavioral blacklisting can apply not only to email traffic, but also to the network-level behavior of hosting infrastructure for scam or phishing attacks. First, I will present a brief overview of our study of the network-level behavior of spammers. Second, I will describe two behavioral blacklisting algorithms that are based on insights from our study of the network-level behavior of spammers. Finally, I will describe our ongoing work applying similar behavioral detection techniques to detecting both online scam hosting infrastructure and phishing attacks. Bio Nick Feamster is an assistant professor in the College of Computing at Georgia Tech. He received his Ph.D. in Computer science from MIT in 2005, and his S.B. and M.Eng. degrees in Electrical Engineering and Computer Science from MIT in 2000 and 2001, respectively. His research focuses on many aspects of computer networking and networked systems, including the design, measurement, and analysis of network routing protocols, network operations and security, and anonymous communication systems. He recently received the Presidential Early Career Award for Scientists and Engineers (PECASE) for his contributions to cybersecurity, notably spam filtering. His honors include a Sloan Research Fellowship, the NSF CAREER award, the IBM Faculty Fellowship, and award papers at SIGCOMM 2006 (network-level behavior of spammers), the NSDI 2005 conference (fault detection in router configuration), Usenix Security 2002 (circumventing web censorship using Infranet), and Usenix Security 2001 (web cookie analysis).

1 0

Colloquium Speaker-THURSDAY, October 29, 2009
by Michele J. Brown 26 Oct '09

26 Oct '09

Title: Network-Level Spam and Scam Defenses Professor Nick Feamster, Georgia Tech Host: Jennifer Rexford CS 105-Small Auditorium 4:30pm, Thursday This talk introduces a new class of methods called "behavioral blacklisting", which identify spammers based on their network-level behavior. Rather than attempting to blacklist individual spam messages based on what the message contains, behavioral blacklisting classifies a message based on how the message itself was sent (spatial and temporal traffic patterns of the email traffic itself). Behavioral blacklisting tracks the sending behavior of an email sender from a wide variety of vantage points and establishes "fingerprints" that are indicative of spamming behavior. Behavioral blacklisting can apply not only to email traffic, but also to the network-level behavior of hosting infrastructure for scam or phishing attacks. First, I will present a brief overview of our study of the network-level behavior of spammers. Second, I will describe two behavioral blacklisting algorithms that are based on insights from our study of the network-level behavior of spammers. Finally, I will describe our ongoing work applying similar behavioral detection techniques to detecting both online scam hosting infrastructure and phishing attacks. Bio Nick Feamster is an assistant professor in the College of Computing at Georgia Tech. He received his Ph.D. in Computer science from MIT in 2005, and his S.B. and M.Eng. degrees in Electrical Engineering and Computer Science from MIT in 2000 and 2001, respectively. His research focuses on many aspects of computer networking and networked systems, including the design, measurement, and analysis of network routing protocols, network operations and security, and anonymous communication systems. He recently received the Presidential Early Career Award for Scientists and Engineers (PECASE) for his contributions to cybersecurity, notably spam filtering. His honors include a Sloan Research Fellowship, the NSF CAREER award, the IBM Faculty Fellowship, and award papers at SIGCOMM 2006 (network-level behavior of spammers), the NSDI 2005 conference (fault detection in router configuration), Usenix Security 2002 (circumventing web censorship using Infranet), and Usenix Security 2001 (web cookie analysis).

1 0

Talk by Marco Canini (EPFL) on Oct 21 from 11-12
by Eric Keller 21 Oct '09

21 Oct '09

If you would like to speak with Marco, let me know. Title: Avenger: Catching Bugs in Distributed Systems Via Almost-Invariant Inference Who: Marco Canini (EPFL) When: Oct. 21st at 11am Where: CS 302 Abstract: It is notoriously hard to develop dependable distributed systems. This is partly due to the difficulties in foreseeing various corner cases and failure scenarios while the system is running over an asynchronous network. Reasoning about the distributed system invariants is easier than reasoning about the code itself. The invariants can be used for debugging, theorem proving, and runtime enforcement. In this talk, I'll introduce an approach to observe the behavior of a system and automatically infer invariants which reveal the bugs in the current implementation of the system. Using our tool, Avenger, we automatically generate a large number of potentially relevant properties, check them against the traces of system executions, and filter out all but a few properties before reporting them to the developer. Our key insight in filtering is that a good candidate for an invariant is the one that holds in all but a few cases, i.e., an "almost-invariant". Our experimental results with the BGP, RandTree, and Chord implementations demonstrate Avenger's ability to identify the almost-invariants that lead the developer to programming errors. Bio: Marco Canini received the "laurea" degree in Computer Science and Engineering from the University of Genoa, Italy. He holds a Ph.D. degree from the Department of Communications, Computer and Systems Science (DIST) of the University of Genoa. During his Ph.D., he was invited as a visitor to the University of Cambridge, Computer Laboratory. He now is with the Networked Systems Laboratory at EPFL, Switzerland. His research focuses on computer networking with emphasis on rethinking Internet fundamentals to include power awareness and improve Internet's energy efficiency, methods for Internet traffic classification based on application identification, design of network monitoring applications, and graphical visualization of networking data.

1 1

Colloquium Speaker-This Wednesday, October 21st
by Michele J. Brown 19 Oct '09

19 Oct '09

Title: Hidden Grammar: Advances in Data-Driven Models of Language Speaker: Noah Smith, Carnegie Mellon University, http://www.cs.cmu.edu/%7Enasmith Date: Wednesday, October 21, 2009 Time: 4:30 PM Host: David Blei Location: Computer Science Auditorium, CS 105 Abstract: With the field of computational linguistics' empirical revolution of the 1990s came the realization that human intuitions about language are insufficient for accurate and robust natural language technologies. The move from hand-written, rule-based models to data-driven techniques led to huge advances, yet we still leaned on human intuition for constructing annotated linguistic datasets. Despite major advances in this paradigm (some of which we'll discuss in this talk), we now know that, in the wild world of real and diverse linguistic data, natural language technology raised on expert-made annotations remains insufficient for real, robust applications. In this talk we adopt the premise that unsupervised learning will, in the long run, be the way forward for learning computational models of language cheaply. We focus on dependency syntax learning without trees, beginning with the classic EM algorithm and presenting several ways to alter EM for drastically improved performance using crudely represented "knowledge" of linguistic universals. We then present more recent work in the empirical Bayesian paradigm, where we encode our background knowledge as a prior over grammars, applying inference to obtain hidden structure. Of course, "background knowledge" is still human intuition. We argue, however, that by representing this knowledge compactly in a prior distribution--far more compactly than the many decisions made in building treebanks--we can experimentally explore the connection between proposed linguistic universals and unsupervised learning. This talk includes discussion of joint work with Shay Cohen, Dipanjan Das, Jason Eisner, Kevin Gimpel, Andre Martins, and Eric Xing.

1 0