Detecting online auction shilling frauds using supervised learning
作者：
Highlights：
• Proposed approach for fraud detection using machine learning and synthetic data.
• Applied approach to detecting two variations of a fraud called competitive shilling.
• Classification models developed were applied to both synthetic and real data.
• Results over synthetic data show a substantial improvement over previous work.
• Results over commercial data show models can find users with suspicious behaviour.
摘要
•Proposed approach for fraud detection using machine learning and synthetic data.•Applied approach to detecting two variations of a fraud called competitive shilling.•Classification models developed were applied to both synthetic and real data.•Results over synthetic data show a substantial improvement over previous work.•Results over commercial data show models can find users with suspicious behaviour.
论文关键词：Supervised fraud detection,Online auction fraud,Agent-based simulation
论文评审过程：Available online 24 October 2013.
论文官网地址：https://doi.org/10.1016/j.eswa.2013.10.033
