SPEAKER:             Tom Mitchell, President-Elect, American Association for Artificial Intelligence             
                                 Professor of AI and Learning Director, 
                                 Center for Automated Learning and Discovery,                                 
                                 School of Computer ScienceCarnegie Mellon University , email: tom.mitchell@cmu.edu  

DATE:                    Friday, November 12, 1999  

TITLE:                    Extracting Information from the World Wide Web  

ABSTRACT:         Consider the fact that although your computer workstation can now retrieve any of 600,000,000 pages on the World Wide Web, it unfortunately cannot understand their content.  This is, of course, because web pages are written to be understandable to people, not computers. The goal of our research is to automatically extract a very large database of facts that mirror the content of the Web, and that can be manipulated by computer.  If we can achieve this goal, it will enable using the web as a gargantuan data base and knowledge base to support a rich variety of applications.  Our approach is to use machine learning algorithms to train a system to automatically extract information from web hypertext.  For example, in one set of experiments our system was trained to extract descriptions of faculty, students, research projects, and courses from web sites of computer science departments.  It then used these learned extraction routines to build a database containing thousands of new entries by automatically browsing new university web sites.  The system is currently running 24 hours per day, and over the past eight months has built a knowledge base containing over 100,000 assertions, with an accuracy of roughly 70%. This talk will present the machine learning algorithms we have developed to date, along with experimental results suggesting these methods can be quite effective for information extraction in certain domains.  

BIO: Tom M. Mitchell is the Fredkin Professor of Artificial Intelligence and Learning in the School of Computer Science , Carnegie Mellon University .  He is also the Founding Director of CMU's Center for Automated Learning and Discovery, an interdisciplinary center for research on data mining. Mitchell is best known for his research on machine learning, in which he has developed applications such as online calendars that learn their users' scheduling preferences, web browsers that learn to extract information from hypertext, and systems that predict birth risks in new pregnancies based on hospital records of previous pregnancies.  Mitchell is the author of the widely used textbook "Machine Learning" (McGraw Hill, 1997), President-Elect of the American Association for Artificial Intelligence, and a member of the Computer Science and Telecommunications Board of the National Academy of Sciences' National Research Council.  Mitchell received his B.S. degree from the Massachusetts Institute of Technology in 1973, and his Ph.D. in electrical engineering from Stanford University in 1979.

SPEAKER:             Berry DeBruijn, University of Ottawa , Email: debruijn@csi.uottawa.ca  

TOPIC:                   Evaluation of Interactive Information Retrieval  

DATE:                    Friday, November 19, 1999  

PLACE:                  Room 318, MacDonald Hall, University of Ottawa  

ABSTRACT:         Our experiment platform - a full-text information retrieval system - supports Query Expansion (QE) by Relevance Feedback (RF). I.e., the system suggests additional query terms that are typical for relevant documents and a typical for irrelevant or unretrieved documents. Evaluation of QE/RF is slightly problematic. A user study is costly and due to tensions between experiment size, expected variations between users and variations between retrieval tasks, effects are difficult to establish statistically and the results are difficult to generalize. A system study lacks the input of users and disregards the interactive character of the system. A system study with a simulated user will lack the essential credibility. A system study with a "cohort" of simulated users - each one behaving slightly differently - could be a step to solve the evaluation problems. The presentation describes: (1) the place that this new experimental method has among existing methods, (2) the design of the study, (3) the design of the group of simulated users, (4) the results of the experiments, (5) the restrictions and assumptions that apply, (6) examples of possible future studies of the framework of the Intelligent Information Access project. How does it relate to my previous Tamale presentation: a more extensive methodological discussion, more results, further analyses, in-depth interpretation, still no clip-art, still no donuts.

SPEAKER:             Joel Martin, National Research Council, Email: joel.martin@iit.nrc.ca  

TOPIC:                   Design of a Better Question Answering System                    

DATE:                    Friday, December 3, 1999 

PLACE:                  Room 318, MacDonald Hall, University of Ottawa 

ABSTRACT:         A better search-engine would allow you to ask a natural language question and would return an answer instead of 10,000 web pages.  In this talk I will review our current design for question answering and summarize the design of all the other systems that were presented  at TREC-8 in Gaithersburg , MD.   Our system, like almost all the other  systems, does 'passage retrieval' (find a short passage likely to contain the answer) and then scans for answer types that match the question (for a Who question look for a person).  The most obvious deficiencies in our current design are speed and the accuracy of "answer-type" identification.  Other passage retrieval systems can do a search in 5-25 seconds while ours takes minutes and other named-entity system components have greater than 90% accuracy while ours is around 50%.  I will finish the talk with an outline of the design of a better, more robust question answering system. This work was done with Chris Lankester from the University of Ottawa .