Abstract: Put forward a lot of data under the dangerous web data mining algorithm based on self-organizing mapping, through massive amounts of data error between the predicted values and the actual value of the judgment and exclude the interference of data, on this basis, through self-organization feature mapping networks for dangerous web data mining, self-organizing feature map network was introduced and the detailed process of the output layer competition, determine a network, can form mapping will be dangerous web data mining as a self-organizing map network input vector input, the output wins produced the corresponding points on the graph, similar input vector convergence in adjacent areas of the map, distance and the region's victory points corresponding to the input vector is dangerous web data can be judge.The simulation results show that the proposed algorithm for huge amounts of data under the interference of dangerous web data mining, not only has high efficiency, and also has high performance in mining precision.