Mobile Roaming Fraud Detection Based On User Behavior In Case Of Ethio Telecom

Mobile roaming data-internet fraud, committed on visitor networks is a continuedrnchallenge and significant source of revenue losses for telecommunicationsrnsocieties including customers. The actually introduced prevention and detectionrnmechanism have limitations in protection of the service.rnIn this study, we used different data-sets and build roaming mobile data fraudrndetection model. Three supervised machine learning algorithms: Artificial NeuralrnNetwork (ANN), Support Vector Machine (SVM) and J48 decision tree (J48rnDT) where used to build model from each data-set. The model performancernwas computed based on different metrics. The model with merged data-setrn(roaming in and roaming out) achieved better performance and J48 DT is resultedrngreater in accuracy of 99.50, average F1_Score 99.00 and ROC 99.30.rnFor compiled usage behavior exceeds the detection of such fraud, organizationrnbetter to periodically analysis of data rather than waiting for TAP file-userrnusage from visited network in addition to revising roaming agreement.

Get Full Work

Report copyright infringement or plagiarism