A Personal Odyssey in the World of Multi-Agent Research
Theme
Creation and Evolution of MAS/DAI Approaches
MAS/DAI ideas (continued)
Precursor Research
Hearsay-II: Parallel, Cooperating Knowledge-Source Model (1973-1975)
Functional Description of the Speech- Understanding KSs
Key Ideas in Blackboard Architecture
Transition to Distributed AI research ó the importance of serendipity
Distributed Hearsay-II Experiments (Erman, 1977-1979)
Network of Hearsay-II Systems
Results of Experiments
Lessons ó It worked but...!
Key Question that Focused Research in the 1980ís
DVMT: an Environment for Research in Cooperative Distributed Problem Solving
DVMT: Agent Architecture (Corkill, 1983)
Integrating Data and Goal-Directed Control and Organizational Structuring
Limitations of Static Meta-Level control (1987)
Partial Global Planning Architecture
Partial Global Planning (Durfee)
Some Important Digressions in Local Agent Control
Further Work on Local Control
Digressions into the World of Negotiation
Key Questions that Focused Research in the 1990ís
Generalizing PGP (Decker)
TAEMS: A Domain Independent Framework for Modeling User Activities
TAEMS (continued)
Information Gathering Example
PPT Slide
Important Aspects of TAEMS
WARREN Style Model of Multi-Agent Information Gathering
GPGP Agent Architecture
Clear Separation of Local Control from Coordination
Integrating GPGP with other Approaches
Putting it all together ó An architecture for Large Agent Societies
Personal Perspective on MAS based on this research path
Personal Perspective (continued)
Current Research Projects
Current Research Projects, Contíd
Questions on my mind
Questions on my mind (continued)
MAS in the 21st Centuryó A Dominant Model
MAS in the 21st Century, Contíd
Important Directions for the Field to Realize this Vision
Directions (continued)
Parting Thoughts
Email: lesser@cs.umass.edu
Home Page: http://mas.cs.umass.edu/
Download presentation source