High communication efforts and poor problem solving results due to restricted overview are two central issues in collaborative problem solving. This work addresses these issues by introducing the processes of agent melting and agent splitting that enable individual problem solving agents to continually and autonomously reconfigure and adapt themselves to the particular problem to be solved.
The author provides a sound theoretical foundation of collaborative problem solving itself and introduces various new design concepts and techniques to improve its quality and efficiency, such as the multi-phase agreement finding protocol for external problem solving, the composable belief-desire-intention agent architecture, and the distribution-aware constraint specification architecture for internal problem solving.
The practical relevance and applicability of the concepts and techniques provided are demonstrated by using medical appointment scheduling as a case study.
Author(s): Markus Hannebauer (eds.)
Series: Lecture Notes in Computer Science 2427 : Lecture Notes in Artificial Intelligence
Edition: 1
Publisher: Springer-Verlag Berlin Heidelberg
Year: 2002
Language: English
Pages: 290
City: Berlin; New York
Tags: Artificial Intelligence (incl. Robotics); Computer Communication Networks; Programming Languages, Compilers, Interpreters; Information Systems Applications (incl.Internet); Computers and Society; Business Information Systems
1.Overview....Pages 3-8
2. Basics of Collaborative Problem Solving....Pages 9-24
3. Distributed Constraint Problems — A Model for Collaborative Problem Solving....Pages 27-57
4. Autonomous Dynamic Reconfiguration — Improving Collaborative Problem Solving....Pages 59-92
5. Multi-agent System Infrastructure....Pages 95-113
6. External Constraint Problem Solving....Pages 115-137
7. Composable BDI Agents....Pages 139-165
8. Internal Constraint Problem Solving....Pages 167-189
9. Controlling Agent Melting and Agent Splitting....Pages 191-213
10. Evaluation....Pages 217-233
11. Conclusion and Future Work....Pages 235-238
A. Symbols and Abbreviations....Pages 241-247
B. An XML-Encoded Request Message....Pages 249-249
C. SICStus Prolog Code for Internal Constraint Problem Solving....Pages 251-258
D. Initialization of the Hospital Scenario Generator....Pages 259-264
