Microsoft Research Silicon Valley Distinguished Speaker Series
2009
Google Fellow Google Inc Amit Singhal
Talk date: 12 October 2009, 3:00-4:30 pm
(refreshments at 2:30pm)
Talk title: Web Search: Challenges and Directions
Talk abstract: These are exciting times for the field of Web search. Search engines are used by millions of people every day, and the number is growing rapidly. This growth poses unique challenges for search engines: they need to operate at unprecedented scales while satisfying an incredible diversity of information needs. Furthermore, user expectations have expanded considerably, moving from "give me what I said" to "give me what I want". Finally, with the lure of billions of dollars of commerce guided by search engines, we have entered a new world of "Adversarial Information Retrieval". This talk will show that the world of algorithm and system design for commercial search engines can be described by two of Murphy's Laws: a) If anything can go wrong, it will; and b) Even if nothing can go wrong, it will anyway.
About the speaker: Amit Singhal is a Google Fellow. According to the New York Times, Mr. Singhal is the master of what Google calls its “ranking algorithm” — the formulas that decide which Web pages best answer each user’s question. A native of India, Amit got his bachelors degree in Computer Science from IIT Roorkee in 1989. Amit holds a MS in Computer Science from University of Minnesota, Duluth, and a Ph.D. in Computer Science from Cornell University in Ithaca, NY. At Cornell Amit studied with Gerard Salton, a pioneer in the field of Information Retrieval. Amit runs a team in Google's Search Quality group. He is and his team are responsible for the Google search algorithms.
Robert E. Dinning Professor of Computer Science and Engineering at the University of Washington
Thomas Anderson
Talk date: 29 September 2009, 3:00-4:30 pm
(refreshments at 2:30pm)
Talk title: A Case for OneSwarm
Talk abstract: OneSwarm is a platform for wide-scale distributed peer-to-peer applications, designed as an open-source alternative to cloud computing for sharing user-generated content. While storing and sharing data through centralized data centers offers many advantages to the system designer, it is not without its drawbacks for the rest of us: a loss of privacy, application lock-in, susceptibility to censorship, applicability to the long tail of unprofitable user-generated content, and limits on system scalability and reliability. Peer-to-peer systems to date suffer from many of these same issues, leaving us without much of an alternative. OneSwarm is our attempt to address these issues, starting with the fundamental assumption of no centralized trust — OneSwarm is not only open source, it is designed to resist being “owned”. This raises questions of incentives, user interface design, storage, privacy, and application management, that will be the focus of the talk.
About the speaker: Thomas Anderson is the Robert E. Dinning Professor of Computer Science and Engineering at the University of Washington. His research interests span all aspects of building practical, robust, and efficient computer systems, including distributed systems, operating systems, computer networks, multiprocessors, and security. He is an ACM Fellow, winner of the ACM SIGOPS Mark Weiser Award, winner of the IEEE Bennett Prize, and he has co-authored over a dozen award papers.
IBM Professor of Engineering and Applied Mathematics in Computer Science, Cornell University
John Hopcroft
Talk date: 17 August 2009, 3:00 pm
Talk title: Future Research Directions in Computer Science
Talk abstract: The field of computer science is changing rapidly due to the increased power of computing, the size and complexity of problems we deal with, the merging of computing and communication, and the availability of vast amounts of information in digital form. Because of this, research in programming languages, compilers, operating systems, and algorithms that has been a major focus of researchers in the past will be supplemented by new areas. In the future we will deal with social networks, tracing the flow of ideas in the scientific literature, high dimensional data, searching, ranking, collaborative filtering and detecting changes in data or transactions before they become obvious. This talk will explore research topics in several of these areas: tracking communities in social networks, tracing the flow of ideas in scientific literature and understanding high dimensional data.
About the speaker: John E. Hopcroft is the IBM Professor of Engineering and Applied Mathematics in Computer Science at Cornell University. He received his BS (1961) from Seattle University and his M.S. (1962) and Ph.D. (1964) in electrical engineering from Stanford University. His research centers on theoretical aspects of computer science. He served as dean of Cornell University’s College of Engineering from 1994 until 2001. He is a member of the National Academy of Sciences, of the National Academy of Engineering, and a fellow of the American Academy of Arts and Sciences, the American Association for the Advancement of Science, the Institute of Electrical and Electronics Engineers, the Association of Computing Machinery, and the Society of Industrial and Applied Mathematics. In 1986 he was awarded the A. M. Turing Award for his research contributions. In 1992, he was appointed by President Bush to the National Science Board, which oversees the National Science Foundation, and served through May 1998. He received the IEEE Harry Goode Memorial Award in 2005, the Computing Research Association’s Distinguished Service Award in 2007, and the ACM Karl V. Karlstrom Outstanding Educator Award in 2009. He has honorary degrees from Seattle University, the National College of Ireland, the University of Sydney and is an honorary professor of the Beijing Institute of Technology. He serves on the Packard Foundation’s Science Advisory Board, Microsoft Technical advisory board for Research Asia and the advisory boards of IIIT Delhi and the College of Engineering at Seattle University.
Dougald Jackson Professor of Electrical Engineering and Computer Science, MIT
Silvio Micali
Talk date: 27 May 2009, 3:00 pm
Talk title: A New Approach to Auctions and Mechanism Design
Talk abstract: Mechanism design's goal is to guarantee a given property P, defined on "state information" known only to a set of players, by leveraging the players' knowledge and rationality. Traditionally, this goal is interpreted as designing a game G such that P holds at one or more of G's equilibria. But due to collusion, as well as computational complexity and privacy concerns, traditional mechanisms may be very far from guaranteeing their desired properties.
We thus put forward a new approach to mechanism design that (1) does not rely on equilibria, (2) is resilient to collusion, and (3) harmonizes incentives with computational complexity and privacy. We exemplify our approach for revenue in combinatorial auctions.
(Joint work with Jing Chen, and part of a joint research effort with Paul Valiant)
About the speaker: Silvio Micali received his Laurea in Mathematics from the University of Rome in 1978, and his Ph.D. in Computer Science form the University of California at Berkeley in 1983. He joined MIT's faculty in 1983, where he is now the Dougald Jackson Professor of Electrical Engineering and Computer science.
Silvio's research interests are cryptography, proof systems, zero knowledge, pseudo-random generation, secure protocols, and mechanism design.
Silvio is the recipient of the Goedel Prize (in theoretical computer science) and the RSA prize (in Cryptography). He is a member of the National Academy of Sciences, the National Academy of Engineering, and the American Academy of Arts and Sciences.
Past Speakers
Professor of Computer Science at Stanford UniversityDaphne Koller
Talk date: 12 November 2008, 3:00pm
Talk title: Probabilistic Models for Holistic Scene Understanding
Talk abstract: Over recent years, computer vision has made great strides towards annotating parts of an image with symbolic labels, such as object categories (things) or segment types (stuff). However, we are still far from the goal of providing a semantic description of an image, such as "a man, walking a dog on a sidewalk, carrying a backpack".
In this talk, I will describe some projects we have done that attempt to use probabilistic models to move us closer towards the goal. The first part of the talk will present methods that use a more holistic scene analysis to improve our performance at core tasks such as object detection, segmentation, or 3D reconstruction.
The second part of the talk will focus on finer-grained modeling of object shape, so as to allow us to annotate images with descriptive labels related
to the object shape, pose, or activity (e.g., is a cheetah running or standing). These vision tasks rely on novel algorithms for core problems in machine learning and probabilistic models, such as efficient algorithms for probabilistic correspondence, transfer learning across related object classes for learning from sparse data, and more.
Dr. Daphne Koller is a professor of Computer Science at Stanford University. Her research focuses on developing and using machine learning and probabilistic methods to model and analyze complex domains, such as biological systems or the physical world. She is the author of over 150 refereed papers in venues as diverse as Science, Nature Genetics, Machine Learning, Neural Information Processing Systems, and Games and Economic Behavior. She is the recipient of numerous awards, including the ONR Young Investigator Award (1998), the Presidential Early Career Award for Scientists and Engineers (1999), the IJCAI Computers and Thought Award (2001), the Cox Medal for Fostering Excellence in Undergraduate Research (2003), the MacArthur Fellowship (2004), and the ACM/Infosys Award (2008).
Adjunct Professor of Computer Science at M.I.TMichael Stonebraker
Talk date: 2 October 2008, 3:00pm
Talk title: Morpheus: A Deep Web Search Engine
Talk abstract: Much attention has been paid to searching the shallow web for information visible on pages accessible to crawlers. However, information behind web forms and other kinds of interfaces is typically invisible to current crawlers, and cannot be found using current shallow web systems. It is estimated that the deep web is 500 times as large as the shallow web. Much of the deep web is data (as opposed to text) and contains transportation schedules, pricing information, availability, tardiness, and personal data.
This talk describes Morpheus, a search system oriented toward the deep web which we have built at M.I.T. The architecture of the systems will be described, and a demo given. Open issues will be considered, and efforts to "wrap" the shallow web will be explored. Finally, some of the experiences of a startup (currently in stealth mode) to commercialize Morpheus will be recounted.
Dr. Michael Stonebraker has been a pioneer of data base research and technology for more than a quarter of a century. He was the main architect of the INGRES relational DBMS, the object-relational DBMS, POSTGRES, and the federated data system, Mariposa. All three prototypes were developed at the University of California at Berkeley where Stonebraker was a Professor of Computer Science for twenty five years. He is the founder of three successful Silicon Valley startups, whose objective was to commercialize these prototypes.
Professor Stonebraker is the author of scores of research papers on data base technology, operating systems and the architecture of system software services. He was awarded the prestigious ACM System Software Award in 1992, for his work on INGRES. Additionally, he was awarded the first annual Innovation award by the ACM SIGMOD special interest group in 1994, and has been recognized by Computer Reseller News as one of the top five software developers of the century. Moreover, Forbes magazine named him one of the 8 innovators driving the Silicon Valley wealth explosion during their 80th anniversary edition in 1998. He was elected to the National Academy of Engineering in 1998 and is presently an Adjunct Professor of Computer Science at M.I.T.
Co-founder and Technical advisor, FAST Enterprise Search; Principal Research Scientist, CS dept, Cornell University Robbert Van Renesse
Talk date: 19 May 19 2008, 3:00pm
Talk title: Building scalable and fault-tolerant enterprise search platforms
Talk abstract: Enterprise search has become a critical part of an organization's infrastructure. At large organizations, documents are generated at a high rate and have to be available for search within seconds. High availability and high performance for search are both essential, while the use of specialized hardware should be avoided. In this talk I will cover some of the distribution and replication techniques that we have developed in order to meet the difficult requirements.
Joint work with Fred B. Schneider, Johannes Gehrke, and Dag Johansen.
Dr. Robbert Van Renesse is a Principal Research Scientist at the Department of Computer Science. He received his Ph.D. from the Vrije Universiteit in Amsterdam in 1989 where he developed the Amoeba Distributed Operating System. Subsequently he worked on the Plan 9 operating system at AT&T Bell Laboratories. Since joining Cornell in 1991 he has worked on fault-tolerant distributed systems. He co-founded D.A.G. Labs that was acquired by FAST, and Reliable Network Solutions whose technology was acquired by Amazon.com. Other companies that use technology developed by Van Renesse include Microsoft, IBM, and Stratus.
FAST is a global provider of enterprise search technologies. FAST's solutions are used by more than 2,600 global customers and partners, including America Online, Dell, IBM, Reuters, and the US Army. FAST is headquartered in Norway. The FAST Group operates globally with presence in Europe, the United States, Asia Pacific, Australia, South America, and the Middle East. For further information about FAST, please visit www.fastsearch.com.
Distinguished Professor, University of Massachusetts Amherst; Director of the Center for Intelligent Information RetrievalBruce Croft
Talk date: 9 June 2008, 3:00pm
Talk title: Longer Queries, Better Answers?
Talk Abstract: Web search engines produce effective rankings for queries consisting of a small number of keywords, within the accepted limitations of what it means for a Web page to be “relevant” to a query. On the other hand, people are perfectly capable of describing what they are looking for more precisely with a longer query. The problem is that we currently don’t know what to do with these longer queries, unless they happen to be a “factoid” query of the type used in question answering systems. Even in TREC evaluations, where query response times are not an issue, long queries generally are less effective than short queries. In this talk, I will review the approaches that have been taken with longer queries, and present two pieces of our research related to this issue; generating keyword queries from long queries and finding answers in a community-based question and answer archive.
W. Bruce Croft is a Distinguished Professor in the Department of Computer Science at the University of Massachusetts, Amherst, which he joined in 1979. In 1992, he became the Director of the Center for Intelligent Information Retrieval (CIIR), which combines basic research with technology transfer to a variety of government and industry partners. He has published more than 180 articles related to information retrieval. Dr. Croft was elected a Fellow of ACM in 1997, received the Research Award from the American Society for Information Science and Technology in 2000, and received the Gerard Salton Award from the ACM Special Interest Group in Information Retrieval (SIGIR) in 2003.
2007 Speakers
Digital Librarian, Director and Co-Founder, Internet Archive.Brewster Kahle
Talk date: 27 March 2007, SVC-1/Jupiter, 12:00pm
Talk title: Universal Access to Human Knowledge (Or Public Access to Digital Materials)
Talk abstract: The goal of universal access to our cultural heritage is within our grasp. With current digital technology we can build comprehensive collections, and with digital networks we can make these available to students and scholars all over the world. The current challenge is establishing the roles, rights, and responsibilities of our libraries and archives in providing public access to this information. With these roles defined, our institutions will help fulfill this epic opportunity of our digital age.
Brewster has built technologies, companies, and institutions to advance the goal of universal access to all knowledge. He currently oversees the non-profit Internet Archive as founder and Digital Librarian, which is now one of the largest digital archives in the world. As a digital archivist, Brewster has been active in technology, business, and law.
Andrew and Debra Rachleff professor and chairman, Dept. of Computer and Information Science, U Penn Fernando Pereira
Talk date: 11 April 2007
Talk title: Learning to Analyze Sequences.
Talk abstract: Sequential data -- speech, text, genomic sequences -- floods our storage servers. Much useful information in these data is carried by implicit structure: phonemes and prosody in speech, syntactic structure in text, genes and regulatory elements in genomic sequences. Over the last six years, several of us have been investigating structured linear models, a unified discriminative learning approach to sequence analysis problems. I will review the approach and illustrate it with applications to parsing, information extraction, and gene finding. I will conclude with a summary of other applications and current research questions.
Joint work with Axel Bernal, Koby Crammer, John Lafferty, Andrew McCallum, Ryan McDonald, and Fei Sha.
Professor Fernando Pereira is chairman of the department of Computer and Information Science, University of Pennsylvania. He received a Ph.D. in Artificial Intelligence from the University of Edinburgh in 1982. Before joining Penn, he held industrial research and management positions at SRI International, at AT&T Labs, where he led the machine learning and information retrieval research department from September 1995 to April 2000, and at WhizBang Labs, a Web information extraction company. His main research areas are computational linguistics and machine learning, and he is a main contributor to several advances in finite-state models for speech and text processing in everyday industrial use. He has 97 research publications on computational linguistics, speech recognition, machine learning, and logic programming, and several issued and pending patents on speech and language processing, and on human-computer interfaces. He was elected Fellow of the American Association for Artificial Intelligence in 1991 for his contributions to computational linguistics and logic programming, and he is a past president of the Association for Computational Linguistics.
Professor, School of Engineering and Computer Science, Hebrew University, and Google. Noam Nisan
Talk date: 26 July 2007 – SVC 1/Saturn, 12:00pm
Talk title: Algorithmic Mechanism Design
Talk abstract: One of the challenges that the Internet raises is the necessity of designing distributed protocols for settings where the participating computers are owned and operated by different owners with different goals. Over the last decade or so there has been much research that aims to address these issues using ideas taken from the micro-economic field of mechanism design. In this talk I will survey the current state of the field: how mechanism design is applied in computational settings, how far can classical ideas go, and what are the challenges for further research. Among the applications discussed will be combinatorial auctions, cost sharing, scheduling, and routing in networks.
Professor Noam Nisan received his Ph.D. from the University of California, at Berkeley, and is now a Professor of Computer Science in the Hebrew University of Jerusalem. He published three books and numerous research papers on algorithms, complexity theory, communication, computerized auctions, and electronic commerce. He has received several professional awards including the ACM 1988 Distinguished Dissertation award for his dissertation "Using Hard Problems to Create Pseudorandom Generators.", and the Michael Bruno award, granted annually by Yad Hanadiv (also known as the Rothschild Foundation) to outstanding Israelis in the field of science and learning, for his research in the field: Electronic markets and auctions and economic mechanisms in computation.
Paul and Marlene Borman Professor, Computer Science and Applied Mathematics, Weizmann Institute Adi Shamir
Talk date: 28 August 28 2007 (SVC-1/Jupiter, 12:00pm).
Talk title: A Top View of Side Channel Attacks
Talk abstract: Side channel attacks are powerful techniques which can bypass the mathematical security of many cryptosystems by observing the physical properties of their implementations. In this talk I will survey some new side channel attacks developed by me and my colleagues during the last couple of years on PC's, smart cards, RFID tags, etc.
Professor Adi Shamir obtained his MSc and PhD in Computer Science from the Weizmann Institute in 1975 and 1977 respectively. His thesis was titled, "Fixed Points of Recursive Programs". After a year postdoc at Warwick University, he did research at MIT from 1977–1980 before returning to be a member of the faculty of Mathematics and Computer Science of the Weizmann Institute. He was one of the inventors of the RSA algorithm (along with Ron Rivest and Len Adleman), and has made numerous contributions to the fields of cryptography and computer science.
Shamir is the winner of the 2002 ACM Turing award, jointly with Leonard M. Adleman, Ronald R. Rivest, "For their ingenious contribution for making public-key cryptography useful in practice." Shamir has also received ACM's Kannelakis Award, the Erdős Prize of the Israel Mathematical Society, the IEEE's W.R.G. Baker Prize, the UAP Scientific Prize, The Vatican's PIUS XI Gold Medal and the IEEE Koji Kobayashi Computers and Communications Award.
professor, School of Computer Science, CMU, and Vice President, Research Akamai Technologies.Bruce Maggs
Talk date: 23 October 2007, SVC-1/Jupiter, 12:00pm
Talk title: Lessons in Engineering Self-Managed Networks
Dr. Bruce Maggs received the S.B., S.M., and Ph.D. degrees in computer science from the Massachusetts Institute of Technology in 1985, 1986, and 1989, respectively. His advisor was Charles Leiserson. In 1994, he joined Carnegie Mellon, where he is now a Professor in the Computer Science Department. While on a two-year leave-of-absence from Carnegie Mellon, Maggs helped to launch Akamai Technologies, serving as its Vice President for Research and Development, before returning to Carnegie Mellon. He retains a part-time role at Akamai as Vice President for Research. Maggs is spending the 2007-2008 academic year at Duke University. He has also held visiting faculty positions at M.I.T. and Princeton University.
Maggs's research focuses on networks for parallel and distributed computing systems. In 1986, he became the first winner (with Charles
Leiserson) of the Daniel L. Slotnick Award for Most Original Paper at the International Conference on Parallel Processing, and in 1994 he received an NSF National Young Investigator Award. He was co-chair of the 1993-1994 DIMACS Special Year on Massively Parallel Computation.
Maggs serves on the ACM Council as a Member-at-Large, and and has served on the steering committees for the ACM Symposium on Parallel Algorithms and Architectures (SPAA) and ACM Internet Measurement Conference (IMC).
Mozah Bint Nasser professor, School of Computer Science, CMURaj Reddy
Talk date: 3 Dec 2007, SVC-6/Titan, 12:00pm
Talk Title: Global Access to Information: Research Issues in Data Mining and Text Mining
Talk Abstract: In this talk we will present research issues that arise when attempting to provide "Global Access to Information". To be true to this vision, one must also resolve the problems of Language Divide and Literacy Divide. Over 80% of the global population is not English-literate and over 20% of the population is functionally illiterate, i.e., they cannot read and understand in any language! We will use example from Million Book Digital Library projectand from a project in India to provide health information to illiterate people. Jaime Carbonell stated, about 10 years ago, that the CMU Language Technology Institute research mission is "getting the right information, to the right people, at the right time, on the right medium, in the right language and with the right level of detail". In spite of major advances in search technologies, we are not close to achieving the information society bill of rights in providing global access to iformation. This talk will provide a forum for discussion on the research agenda in Data Mining and Text Mining necessary for fulfilling this vision.
Dr. Raj Reddy began his academic career as an Assistant Professor at Stanford in 1966. He has been a member of the Carnegie Mellon faculty since 1969. He served as the founding Director of the Robotics Institute from 1979 to 1991 and the Dean of School of Computer Science from 1991 to 1999. Dr. Reddy's research interests include the study of human-computer interaction and artificial intelligence. His current research interests include Million Book Digital Library Project; a Multifunction Information Appliance that can be used by the uneducated; Fiber To The Village Project; Mobile Autonomous Robots; and Learning by Doing.
He is a member of the National Academy of Engineering and the American Academy of Arts and Sciences. He was president of the American Association for Artificial Intelligence from 1987 to 89. Dr. Reddy was awarded the Legion of Honor by President Mitterand of France in 1984. He was awarded the ACM Turing Award in 1994, the Okawa Prize in 2004, the Honda Prize in 2005, and the Vannevar Bush Award in 2006. He served as co-chair of the President's Information Technology Advisory Committee (PITAC) from 1999 to 2001 under Presidents Clinton and Bush.
Practical Information
Talks are open to the public. Events will be held in the Titan conference room in Building 6 on Microsoft’s Silicon Valley Campus (1288 Pear Avenue), and light refreshments will be served. E-mail will be sent to lab members a few weeks before each appearance.
Series Coordinators
- Marcos K. Aguilera, Microsoft Research
- Ariel Fuxman, Search Labs
