MACC'97 Session: Learning, Evolutionary Computing
In this paper, we propose to apply the multi-agent model which consists of the heterogeneous agents to the image retrieval support. In this model, there are two species agents; the information filtering agent and the information discovery agents. This support mechanism is implemented by creating an artificial ecosystem with the evolving information filtering agents and information discovery agents that cooperate and compete in an information retrieval. This model can adapt various and many users by using machine learning method, and can adapt the complex retrieval process by using the communication between information filtering agents and discovery agents.
Image Retrieval, Retrieval Support, Communication, Self-organization
PS
file(+gzip) (in Japanese)