Name
Abstract | ||
|---|---|---|
The online shopping sector is continuously growing, generating a turnover of billions of dollars each year. Unfortunately, this growth in popularity is not limited to regular customers: Organized crime targeting online shops has considerably evolved in the past years, causing significant financial losses to the merchants. As criminals often use similar strategies among different merchants, sharing information about fraud patterns could help mitigate the success of these malicious activities. In practice, however, the sharing of data is difficult, since shops are often competitors or have to follow strict privacy laws. In this paper, we propose a novel method for fraud detection that allows merchants to exchange information on recent fraud incidents without exposing customer data. To this end, our method pseudonymizes orders on the client-side before sending them to a central service for analysis. Although the service cannot access individual features of these orders, it is able to infer fraudulent patterns using machine learning techniques. We examine the capabilities of this approach and measure its impact on the overall detection performance on a dataset of more than 1.5 million orders from a large European online fashion retailer. |
Year | DOI | Venue |
|---|---|---|
2018 | 10.1007/978-3-030-01701-9_22 | SecureComm (1) |
Field | DocType | Citations |
Bloom filter,Internet privacy,Popularity,Organised crime,Privacy laws of the United States,Competitor analysis,Business | Conference | 0 |
PageRank | References | Authors |
0.34 | 24 | 5 |
Name | Order | Citations | PageRank |
|---|---|---|---|
| Daniel Arp | 1 | 46 | 3.44 |
| Erwin Quiring | 2 | 15 | 4.49 |
| Tammo Krueger | 3 | 164 | 9.72 |
| Stanimir Dragiev | 4 | 0 | 0.34 |
| Konrad Rieck | 5 | 1585 | 85.84 |