叶艺勇



摘 要: 为了降低电子商务交易的风险,需要进行交易风险的量化评估,提出一种基于改进遗传算法的电子商务交易风险评估方法。采用稀疏散点云数据采集技术进行电子商务交易信息的数据采样,并输入到云存储系统中建立电子商务交易风险数据评估的专家数据库,提取电子商务平台中商家的信任度推荐参量值,并进行信息融合处理。对融合后的商家信任度信息采用遗传算法进行交叉变异处理,结合自适应全局遗传进化方法实现电子商务交易风险信息的准确预测,从而实现交易风险评估。仿真结果表明,采用该方法进行电子商务交易风险评估的预测准确性较好,收敛误差较低,具有可行性。
关键词: 遗传算法; 电子商务; 交易; 信息融合; 风险评估
中图分类号: TN99?34; TP391 文献标识码: A 文章编号: 1004?373X(2017)13?0094?04
Abstract: In order to reduce the risk of e?commerce transaction, it is necessary to perform the quantitative evaluation of transaction risk, therefore an e?commerce transaction risk evaluation method based on improved genetic algorithm is put forward. The sparse scattered point cloud data technology is used to sample the data of the e?commerce transaction information, and input it into the cloud storage system. The specialist database of the e?commerce transaction risk data evaluation was established. The merchant trust recommendation parameters in e?commerce platform are extracted, and conducted with information fusion. The genetic algorithm is used to perform the crossover and mutation for the fused merchant trust information, and combined with the adaptive global genetic evolution algorithm to predict the e?commerce transaction risk information accurately, and realize the transaction risk assessment. The simulation results show that the method has high prediction accuracy and low convergence error for e?commerce transaction risk assessment, and is feasible.
Keywords: genetic algorithm; e?commerce; transaction; information fusion; risk assessment
0 引 言
网络技术和现代物流技术催生了电子商务的快速发展,电子商务平台建立在P2P和O2O交易平台基础上,交易平台具有开放性和自组织性,导致电子商务交易的管理和控制有漏洞,容易出现交易风险,为商家和顾客带来了较大的损失[1]。
为了最大限度地降低电子商务交易的风险,需要采用量化信息评估方法进行电子商务交易预测评估,提高应对风险的能力和水平,因此,研究电子商务交易风险评估方法具有重要意义。
传统方法主要采用神经网络预测评估方法和决策树预测评估方法进行电子商务交易风险评估,采用无……