High-level Behavior Representation Languages Revisited
Abstract
There has only been a short history of high level languages to model human cognition based on cognitive architectures. TAQL is an early example (Yost, 1993). TAQL showed a large (3x) speed increase over plain Soar, but because it did not support Soar's learning mechanism and because Soar changed soon after its release, TAQL's impact was not as great as its developer probably would have liked. It is time again to consider high level behavior representation languages. Cognitive models and intelligent agents are becoming more complex and pervasive. This is driving the need for development environments that make it easier to create, share, and reuse cognitive models. Several high level modeling languages have recently been created and several of them are described briefly here. These languages are each different, but they have a common goal of making modeling human data easier to perform. We can now see some generalities and common lessons. By holding this symposium we will identify lessons for the development of these languages as well as for their users. These languages are reviewed briefly in the next section.
Document Details
- Document Type
- Technical Report
- Publication Date
- Feb 28, 2006
- Accession Number
- ADA631086
Entities
People
- Andrew Howes
- Bonnie John
- Brad Best
- Christian Lebiere
- Frank Ritter
- Jacob Crossman
- Mark Cohen
- Randolph M. Jones
- Richard L. Lewis
- Steven R. Haynes
Organizations
- Pennsylvania State University