LAUSR

Abstract Case revision has always been challenging in case-based reasoning (CBR) processes. Most CBR methods used for analyzing geological disasters fail to consider the spatial relationships among geological environmental factors. Therefore, conventional case revision rules do not allow for effective case-based reasoning for geologic disaster assessment. In this study, we first establish a spatial case library of historical geological disasters. Subsequently, these spatial cases are organized, reduced, and weighted using spatial clustering, genetic algorithm and rough set hybrid algorithms, respectively. Based on these efforts, we propose a retrieval method for the spatial cases and evaluate the proposed solution for a target case. Finally, a geological disaster assessment model possessing a spatial case revision function is obtained by combining geographic information system (GIS) technology with the genetic algorithm. Experiments and applications show that our spatial revision approach is effective and that it exhibits a higher classification performance compared to traditional data mining methods with regard to geological disaster assessment. The approach also exhibits a higher accuracy and efficiency than typical spatial CBR or case revision models do. The results of this study thus facilitate rapid geological disaster assessment, making it more facile and convenient for decision makers to execute decisions efficiently and quickly.