An intelligent data mining and searching agent based on watermarking technology

In this paper, we present a novel technique of intelligent agent, which can perform digital watermarking scheme on variant file types such as jpeg, bmp, wav, mpeg, etc. "Watermarks" embedded in are XML self-describing information of corresponding files, which are invisible and can be transferred with their related files when they are copied or moved to different places. An agent can "understand" a file by extracting the self-describing information embedded in it. With such information some critical requirements can be satisfied, like finding a picture with more than three beautiful girls or calculating how many songs sing by Chinese children existed on local network. Finally, scalability and efficiency of the agent are discussed.