关键词: 数据挖掘, 案例推理, 修复技术, 熵权-逼近理想解排序法

Abstract: The screening of remediation technologies is a crucial process for efficiently repairing contaminated sites and improving land reuse. Currently, the screening index system for site remediation technology is too complex, and the weights of indicators rely on expert judgment. To overcome these challenges, we examine site characteristics and the implementation effects of remediation technologies with 514 Chinese cases from 2010 to 2022. A screening index system was established. The cases were divided into groups according to the remediation technology applied, and the most similar case to the specific site within each technical group was selected by the K-nearest neighbor method (KNN). This allowed for a comparison of implementation effects to quantify the indices associated with each technology. The entropy-TOPSIS method was then applied to evaluate the effectiveness of different remediation technologies. Taking a retired non-ferrous metal smelter site in Northeast China as an example, the most suitable remediation technology that aligned with its actual application was successfully selected. In conclusion, this method demonstrates high practicality and feasibility, which would be helpful for environmental protection workers and enterprises in China.