Privacy-Preserving Fraud Detection via Cooperative Mobile Carriers with Improved Accuracy

• Main Authors: Wenyan Yao, Na Ruan, Feifan Yu, Weijia Jia, Haojin Zhu
• Format: Conference Proceeding
• Language: English
• Subjects: Data models; Data privacy; Mobile communication; Privacy; Resource management; Security
• ISSN: 2155-5494
• DOI: 10.1109/SAHCN.2017.7964943

With the explosive growth of users in mobile carrier, telecommunication fraud causes a serious loss to both of the users and carriers. The academia has an increasing interest in the issue of detecting and recognizing fraudster, and varies strategies have been proposed to prevent the attack and fraudulent activity. However, fraudsters are always inclined to hide their identity and perform the fraudulent activity through different mobile carriers, which makes the previous methods less effective in fraud detection. In this paper, we propose a novel strategy with a high accuracy and security through the cooperation among mobile carriers. We introduce the Latent Dirichlet Allocation (LDA) model to profile users in different carriers. In order to match the fraud accounts, we propose a strategy based on Maximum Mean Discrepancy (MMD) to analyze and compare the distribution of statistical samples. Meantime, during the cooperation of carriers, there is a risk of privacy disclosure. To deal with this weakness, we also demonstrate that our method can detect the fraudulent accounts without leaking the private records and data of user accounts based on the differential privacy.