Detection of Feedback Reputation Fraud in Taobao Using Social Network Theory

As the number of shoppers in C2C sites is rapidly growing in China, online customers may suffer from online fraud. The feedback reputation system has been used to prevent abusing behavior and secure transaction. However, large numbers of fraudsters manipulated their feedback scores by engaging in a number of legitimate sales. Furthermore, there exist agents of "Credit speculation", who boost the fraudsters\´ scores professionally in China. Since fraud behaviors are subtle and complex. it makes the fraudulent behavior more difficult to discover. In this study, we seek for extracting characteristic features in fraudsters\´ behaviors. We present an approach for analyzing online crediting behaviors, which utilizes the social network analysis (SNA) technique to detect the relationship of potential fraudsters. Some cases of real data from Taobao demonstrate our analyses.