Turing Center Distinguished Lecture

Thomas Wasow (Linguistics and Philosophy, Stanford University)
Selecting Among Paraphrases
November 4 (Friday), 3:30-5:00 pm, Mary Gates Hall 241
Recent publications: "The Puzzle of Ambiguity", "Intuitions in Linguistic Argumentation"

Turing Center Distinguished Lecture

Lenhart Schubert (Computer Science, University of Rochester)
Turing's Dream and the Knowledge Challenge
November 10 (Thursday), 3:30-5:00 pm, Electrical Engineering 105
See and hear this lecture.
Recent publications: "Can We Derive General World Knowledge from Texts?", "Some KR&R Requirements for Self-Awareness"

Discussion

Philip N. Howard, Ibrahim Al Beayeyz, Hans Jochen Scholl
Dialogue on the Impact of Communication and Information Technology in Global Contexts
November 15 (Tuesday), 10:00-11:00 am, Mary Gates Hall 420

Turing Center Reception

Introduction to the Turing Center
Reception, remarks, and open discussion to introduce the Turing Center and learn about related work being done elsewhere on campus and in industry. Light lunch buffet will be served.
November 15 (Tuesday), 12:00 noon to 2:00 pm, Paul G. Allen Center for Computer Science and Engineering 691 (Bill and Melinda Gates Commons)

Turing Center Distinguished Lecture

Alon Lavie (Language Technologies and Computer Science, Carnegie Mellon University)
Learning-Based Approaches for Developing MT [Machine Translation] Prototypes for Languages with Limited Resources
November 17 (Thursday), 3:30-5:00 pm, Electrical Engineering 105
See and hear this lecture.
Recent publications: "Multi-Engine Machine Translation Guided by Explicit Word Matching", "MT for Minority Languages Using Elicitation-Based Learning of Syntactic Transfer Rules"

Turing Center Distinguished Lecture

Norbert Fuchs (Informatics, University of Zurich)
Knowledge Representation and Reasoning in (Controlled) Natural Language
December 6 (Tuesday), 3:30-5:00 pm, Electrical Engineering 105

See and hear this lecture.

Related publications:

Turing Talk

Noah Friedland
Project Möbius: Learning by Reading?
December 8 (Thursday), 2:00-3:00 pm, Paul G. Allen Center for Computer Science and Engineering 403
Recent publications: "Project Halo: Towards a Digital Aristotle", "Towards a Quantitative, Platform Independent Analysis of Knowledge Systems"

Abstract

The vast plurality of human knowledge--science, history, philosophy and literature--is captured in written documents. Computer science is only at the beginning of the long journey to make these materials comprehensible to machines. DARPA has funded a 9-month pilot effort to determine whether significant progress can be made in this area. Project Möbius, which began on November 15, contemplates a notional architecture comprised of five elements: (1) a knowledge repository, where harvested knowledge is gradually assembled; (2) a knowledge-driven knowledge acquisition component, capable of recovering a variety of knowledge types from text; (3) a knowledge integration component, which considers how harvested knowledge elements might be integrated into the expanding knowledge repository; (4) a test generation component, to verify that the integrated knowledge is capable of performing problem-solving tasks; and (5) an introspective component, capable of both amending the knowledge repository and instructing the system to acquire new knowledge when tests fail. During the pilot, the research team will attempt to produce new ideas on how text-based knowledge might be represented, integrated, debugged and reasoned over.

Speaker

Noah Friedland was educated in aeronautical and electrical engineering at the Technion in Haifa, Israel, and received a Ph.D. in computer science from the University of Maryland at College Park. He has held engineering positions at Lockheed Martin Astronautics, the University of Washington, and Asta Networks, and managerial positions at LiveBid.com, Tahajo Technologies, and Vulcan, Inc. His work has made contributions to agent-collaborative image processing/understanding, pattern recognition, data mining, computer vision, real-time transit-vehicle location, transit-system performance reporting, massively scalable live-auction and collaborative e-commerce application architectures, distributed bandwidth management, router policy performance testing, network security, and knowledge representation and reasoning. His most recent major project was Project Halo, an effort to develop an artificial intelligence platform of applications to support science and education. From 2002 until 2004 he served as the Program Manager for this project at Vulcan, Inc.

Turing Talk

Wanda Pratt (Information School and Biomedical & Health Informatics)
Interaction Design for Literature-Based Discovery
December 9 (Friday), 1:30-2:20 pm, Paul G. Allen Center for Computer Science and Engineering 303
Recent publications: "LitLinker: Capturing Connections across the Biomedical Literature", "Collaborative Information Synthesis", "Better Rules, Fewer Features: A Semantic Approach to Selecting Features from Text"

Abstract

Because huge quantities of information are published each day, most researchers struggle to keep abreast of work within their own narrow specialization and spend little or no time examining the literature from other related disciplines. However, such isolation can inhibit research progress; many innovations occur only when traditional field boundaries are bridged. We have developed a system, called LitLinker, which provides a first step toward meeting this need. LitLinker bridges traditional field boundaries to identify and link together previously obscured connections in the biomedical literature. We have experimented with various knowledge-based methodologies, natural-language processing techniques, and data-mining algorithms to mine the biomedical literature for new, potentially causal links between biomedical terms. A key component of our approach is the interface that allows researchers to explore the identified connections interactively, and the design of this component will be the main focus of this presentation.

Speaker

Wanda Pratt is an Associate Professor in both the Information School and the Division of Biomedical & Health Informatics in the Medical School at the University of Washington. She is also the Chair of the Ph.D. Program in Information Science at the University of Washington. In 1999, she received her Ph.D. in Medical Informatics from Stanford University. She received her M.S. in Computer Science from the University of Texas, and her B.S. in Electrical and Computer Engineering from the University of Kansas. Her experience in knowledge-based systems dates back to the late eighties when she developed a medical expert system for NASA, and the early to mid nineties when she worked on the CYC project to represent all of common-sense knowledge. Her recent research focuses on knowledge-based methods to retrieve, organize, mine, and present medical information. She is also a recipient of an NSF CAREER Award, which funds this literature-based discovery research.

Turing Talk

Shaojun Wang (Computing Science, Alberta)
Exploiting Syntactic, Semantic and Lexical Regularities in Language Modeling
December 12 (Monday), 11:00 am - 12:00 noon, Paul G. Allen Center for Computer Science and Engineering 403
Recent publications: "Combining Statistical Language Models via the Latent Maximum Entropy Principle", "Exploiting Syntactic, Semantic and Lexical Regularities in Language Modeling via Directed Markov Random Fields"