To provide a data processing system in which a hardware resource can be efficiently utilized, and spatial constraint can be reduced.

Image recognition and synthesis or speech recognition and synthesis devices 40-46 which need a high speed processing or a large-scaled storage capacity are prepared only on a server 102 side. The input and output of images or speeches are operated on a client 100 side. Image data or speech data inputted on the client 100 side are transferred through a network 104 to the server 102 side, and recognized on the server side. Then, the recognized image data and speech data or image data and speech data synthesized on the server side are transferred through the network 104 to the client side, and displayed or reproduced on the client side.