Abstract: During the operation of the Three Gorges Reservoir, the reservoir water level fluctuation could induce old landslide revivification and trigger new landslides (neogenic landslides). The aim of this study is to provide basis for the prevention and treatment of these neogenic landslides. A prediction method with object-oriented C5.0 decision tree model was presented. Evaluation factors were selected, including reservoir water fluctuation, slope gradient, engineering rock group, slope structure and image texture. On the basis of RS interpretation, field investigation, and with RS, GIS and spatial data mining technology, the spatial prediction model for neogenic landslide was built, and the susceptibility map was obtained. The bank slope of Zigui segment in the reservoir area was taken as a case study. Verification shows that the prediction results are consistent with the actual situation. 

Key words: Three Gorges Reservoir area, neogenic landslide, object-oriented, C5.0 decision tree, spatial prediction 