This paper examines the problem of detecting and managing fraud in next-generation telecommunications networks and presents work in progress on implementing a rules-based fraud detection system. The motivations for committing fraud and the different types of fraud that can be perpetrated in next generation networks are examined. The fraud detection system provides an interface that allows users to enter rules that describe possible fraud scenarios. If a rule proves true, an alert is raised indicating that the analyzed record warrants more attention. Features of the system include the use of flexible data formats, spreadsheet based rules specification and the deployment of data reduction techniques to efficiently profile user behavior.
|Title of host publication||Proceedings of SCI 2004|
|Publication status||Published - 2004|
- Fraud Detection
- Next-Generation Networks
- Rules Specification
- Usage Data Analysis