Community Detection for Mobile Money Fraud Detection

This paper presents an overview of a first-year PhD thesis. The main idea can be expressed as a community detection issue in a real network containing data from a telecommunications operator. More precisely, mobile money transactions will be studied, with a view to detecting fraudulent ones. For this purpose, we propose to address the problem by studying different community detection approaches. Indeed, in a network, communities are constituted between users sharing common habits, therefore fraudulent transactions may appear as unusual behaviors. The originality of this work is to process real data, gathered by a telecom operator. Complex network organization is expected, making this thesis quite promising.