A multi-agent framework for meeting scheduling using fuzzy constraints

Based on fuzzy techniques, the paper develops an agent-based approach for meeting scheduling problems. Compared with previous works, it is novel in three aspects. First, a meeting scheduling problem is modeled by fuzzy constraint satisfaction problems in a multi-agent environment. Second, a kind of selfish protocol is presented. An appointment made through this protocol is an overall optimum common time slot. Third, an axiomatic framework is identified for fusing agents' individual evaluations for a common time slot. The framework is also applicable to solution synthesis in distributed expert systems, parallel combination operations in expert systems and aggregation operations in fuzzy mathematics. It is worth further developing: (1) other protocols for more complicated meeting scheduling problems; and (2) other models for fusing agents' individual evaluations for a common time slot.