Abstract: The artical a based on artificial immune algorithm for this potential threat to large e-commerce web database data mining methods.Computing potential threats in web database data attributes,the difference of the potential threat data obtained characteristics of artificial immune to potential threat data,in the process of immune by calculating the matching of antigen and antibody,can obtain potential threat data correlation,combined with the potential threat to potential threat data mining model is set up the correlation of data,has realized the potential threat to large e-commerce web database of data mining.Experimental results show that the improved algorithm to large potential threats e-commerce web database,data mining can improve the accuracy of mining,effect is satisfactory.