Abstract: To efficiently excavate the knowledge from substantial and dynamic landslide monitoring data,we put forward a data mining approach using oracle trigger to monitor data. In order to improve the fitting precision of forecasting model,the time series model ARIMA (Autoregressive Integrated Moving Average Model) was employed to forecast the accumulative displacement and the Oracle trigger was used to refine the monitoring data and optimize the model parameter. Bazimen landslide was taken as a case study. The results indicate that the method improves the mining result of traditional static data and helps people to realize the value of dynamic data in landslide prevention.

Key words: dynamic data mining , landslide monitoring , Oracle trigger , ARIMA