Initial value generation for matchmaking in middle agents

One of the major challenges that agents used in open environments must face is that they must be able to find each other. This is because in an open environment, agents might appear and disappear unpredictably. To address this issue, middle agents have been proposed. The performance of middle agents relies heavily on the matchmaking algorithms used. Matchmaking is the process of finding an appropriate provider for a requester through a middle agent. The practical performance of service provider agents has a significant impact on the matchmaking outcomes of middle agents. Thus the track records of agents in accomplishing similar tasks in the past should be taken into account in matchmaking process. Considering that there are no track records available at the launching of an agent system, this paper discusses some ways to provide reasonable initial values for the track  records. With the agents' history and the initial vallies ofthe track records, the performance of matchmaking algorithms can be improved significantly.