基于协同过滤的思想,提出一种QoS预测算法DRaC,该算法利用聚类技术实现QoS数据的平滑化预处理,同时利用反向预测和交叉预测机制解决数据稀疏问题,提高了预测准确度.最后基于真实的数据集,以NMAE值为标准,验证了DRaC算法相对于协同过滤等传统方法具有更好的预测效果,并衡量了各参数对预测结果的影响. Based on the idea of ​​collaborative filtering, this paper proposes a QoS prediction algorithm, DRaC, which uses the clustering technique to smooth the QoS data and uses the reverse and cross-prediction mechanisms to solve the data sparsity and improve the prediction accuracy. Based on the real data set, the NMAE value is used as a standard to verify that the DRaC algorithm has a better prediction effect than the traditional methods such as collaborative filtering, and the impact of each parameter on the prediction result is also measured.