|
  |
|
|
Krasnow Institute > Monday Seminars > Abstracts A Unified Theory Approach to Modeling Learning in a Time-Pressured Environment Bonnie E. John Almost all computational modeling of human problem-solving and learning in cognitive science has been done on a single task at a time. That is, there have been models of problem solving in the Tower of Hanoi, of learning to program in LISP, of reading comprehension, using text-editors, etc., etc. However, people learn to perform all of these tasks, bringing with them the all that they know from their previous experience. In the 1987 William James Lectures at Harvard University, Allen Newell suggested that cognitive science had arrived at the possibility of creating unified theories that could produce the full range of human cognition, and presented an exemplar: Soar. In this talk, I will summarize the Soar unified theory of cognition and several Soar models of isolated capabilities (e.g., natural-language comprehension, visual search). I will then show how these models can be brought together to model more complex tasks that require the component capabilities and where the unified theory approach will take us next.
The Krasnow Institute for Advanced Study |
|
