文档介绍：Discovery of Fraud Rules for munications -
Challenges and Solutions
Saharon Rosset, Uzi Murad, Einat Neumann, Yizhak Idan, Gadi Pinkas
Amdocs (Israel) Ld.
8 Hapnina St.
Ra’anana 43000, Israel
Email : {saharonr, uzimu, einatn, yizhaki }@
ABSTRACT their carrier or service provider. The Chum Management
application aims at the dual purposesof understanding the driving
Many fraud analysis systems have at their heart a rule-based forces and reasonsbehind chum and predicting the likely future
engine for generating alerts about suspicious behaviors. The rules churners.
in the system are usually based on expert knowledge. Automatic Using “white-box” methods of rule-discovery, the system
rule discovery aims at using past examples of fraudulent and generatesrules or segmentsthat describe the patterns relevant to
legitimate usageto find new patterns and rules to help distinguish chum. These rules can help an analyst understand the reasonsfor
between the two. Some aspectsof the problem of finding rules chum and devise preventive measures. Thus the analyst can
suitable for fraud analysis make this problem unique. bine the automatically generatedknowledge with his domain
them are the following: the need to find bining both the examplesof automatically generatedpatterns (this is
properties of the customer (., credit rating) and