Abstract: An artificial immune response model based on biological immune response mechanism was proposed, which is made up of four procedures: antigen matching, clone selection, mutation and affinity maturation. The diversity of antibodies, achieved by clone and mutation process, can make the built system possess self-adaptive capability. Knowledge acquisition results from affinity maturation. The model needs few operating parameters and is easy to stable. When applied to data analysis the model has fascinating ability to enhance data compression ratio and data recognition ratio.