News Archives

[Colloquium] Data Mining in Word Learning and the Internet Classroom

April 1, 2008

Watch Colloquium: 


  • Date: Tuesday, April 1st, 2008 
  • Time: 11 am — 12:15 pm 
  • Place: ME 218

Paul R. Cohen
Information Sciences Institute University of Southern California

Abstract: The first part of this talk is about a project to have softbots and robots learn the meanings of words. Our goal is to mimic the context in which children learn word meanings, where the learner communicates about what’s going on in the immediate environment with a facilitative, forgiving, capable language user. I will describe recent results from Wubble World, a game environment in which kids and their softbots communicate in English, and the softbots gradually learn word meanings. The second part of the talk is about K12 education and a project we call the internet classroom. I will quickly survey opportunities afforded by the internet classroom for research in many areas of computer science, then focus on the problem of selecting the next task or learning activity in a way that is customized to each student.

Bio: Paul Cohen attended UCSD as an undergraduate and Stanford for his PhD in Computer Science and Psychology. He was a professor at the University of Massachusetts from 1983 – 2003, when he joined the Information Sciences Institute at the University of Southern California. At ISI, he serves as director of the Center for Research on Unexpected Events and as deputy director of the Intelligent Systems Division. At Stanford, Cohen edited the Handbook of Artificial Intelligence with Avron Barr and Edward A. Feigenbaum (extending to four volumes, eventually) which probably explains why he tries hard, though with mixed success, to understand how AI and other cognitive sciences are co-developing. Cohen is especially interested in challenge problems and research methods, and he wrote a book and many articles, and advises government agencies, on these subjects. His research develops new methods for learning, planning, and data mining, and applies them to modeling cognitive development, intelligence analysis, various military problems, and education. Cohen is a fellow of the American Association for Artificial Intelligence.