Pyramid Selling Group Detection Method Based on Community Detection

With the rapid development of the Internet and mobile information technology, there are more and more social network groups. As a malignant group in the social network, the network pyramid selling group makes the pyramid selling crime more harmful and more difficult to track and investigate. In order to more effectively monitor and crack down on online pyramid selling crimes, it is necessary to explore the social activities of online pyramid selling, obtain and analyze its social network data, and quickly and accurately detect and discover online pyramid selling groups in social networks. Based on community detection technology, this paper analyzes the characteristics of pyramid selling in social networks, and proposes a pyramid selling group detection method combining social characteristics. This method sorts out and analyzes the text features and social attributes of pyramid selling, and uses local expansion technology for group detection. This paper also compares this method with several classical community detection algorithms in the application of pyramid selling group detection, which shows the effectiveness of the proposed method. The pyramid selling group detection method combined with social features can achieve an accuracy of 0.929, and is better than other methods in terms of precision, recall rate and F-measure.