B2B平台的反欺诈问题研究

Research on the B2B Platform Anti-fraud Problem

B2B电商平台的欺诈问题一直困扰着电商平台的经营者。以往利用数据挖掘技术解决B2B问题的研究中仍然存在着一些不足。论文利用真实B2B平台公司数据,采用基于决策树(Decision tree)的集成学习算法——Easy-Ensemble对B2B平台反欺诈问题进行研究。实验结果表明,Easy-Ensemble算法的确是解决数据类别不平衡性的一个有效的算法,适用于B2B平台反欺诈问题研究。论文在实验结果的基础上进行深入的商业分析,为B2B企业欺诈问题提出行之有效的建议。

The fraud in B2B business platform has troubled the platform operations. There are still many gaps and deficiencies in data mining technology area about how to solve these frauds. This paper use the real data of a large B2B e-commerce company and chose the ensemble algorithm Easy-Ensemble, which is based on Decision Tree to deal with the problem of B2B fraud. From the experiments results, we can find that Easy-Ensemble algorithm is effective to solve the problem of class imbalance and suit to solve the problem of B2B anti-fraud. As a result, we can provide supports and suggestions for the anti-fraud problems on B2B platform.