Communicating Agents in Prolog

Supervisor: Dr. F. Sadri (Room 447)

The purpose of the project is to extend and adapt an existing implementation in Sicstus Prolog of an agent inference engine based on an abductive proof procedure called the IFF-proof procedure. This inference engine allows reasoning with knowledge containing rules and integrity constraints. The adaptation should allow agents using this inference engine to communicate with each other to ask for and exchange resources in order to obtain the resources they need to satisfy their goals.

To regulate communication between agents, the "rules" of communication (protocols) are often represented as finite state diagrams, using arcs and arrows. These show how an agent should start a dialogue and what it can utter in response to the messages it receives from other agents.

The project involves the following:

The project can be extended in a number of different ways depending on the outcome of the above.