Applying case-based reasoning technique to retaining wall selection

Case-based reasoning (CBR) approach recalls and learns from previous cases to resolve or provide recommendations for current problems. This study presents a case-based retaining wall selection system (CASTLES) in which the case base consists of 254 previous retaining wall cases in design reports in Taiwan. According to the ability of the user's input to accurately describe the characteristics of a new project and a predefined similarity function, CASTLES identifies a set of feasible retaining-wall systems from the case base. Comparing CASTLES with four actual field cases reveals that the case-based reasoning approach is highly promising for selecting retaining wall systems.