Title: Use of fuzzy clustering and support vector machine for detecting fraud in mobile telecommunication networks

Authors: Sharmila Subudhi; Suvasini Panigrahi

Addresses: Department of Computer Science and Engineering, Veer Surendra Sai University of Technology, Burla, Odisha – 768018, India ' Department of Computer Science and Engineering, Veer Surendra Sai University of Technology, Burla, Odisha – 768018, India

Abstract: This paper addresses the problem of finding out fraudulent calls in mobile phones by analysing the user's calling behaviour. In this work, we have used support vector machine (SVM) along with fuzzy clustering for detecting fraudulent usage of mobile phones. The reality mining data-set has been used for testing the efficacy of the proposed approach. A total of five relevant features are being used in creating the user profile from the user's call record. Fuzzy clustering is applied for generating the SVM classifier model. An anomaly is detected when a call pattern does not match with any of the normal patterns. Our experiments show promising results in terms of finding fraudulent calls without raising too many false alarms. Comparative studies are carried out on the proposed system by applying different types of SVMs along with various fuzzy clustering techniques for analysing the performance of the system.

Keywords: mobile communications; mobile telecommunications; mobile networks; fraud detection; feature extraction; support vector machines; SVM; fuzzy clustering; fraudulent calls; mobile phones; cell phones; caller behaviour; user profiles; call records; call patterns; telecommunications networks.

DOI: 10.1504/IJSN.2016.075069

International Journal of Security and Networks, 2016 Vol.11 No.1/2, pp.3 - 11

Received: 04 Nov 2014
Accepted: 17 Apr 2015

Published online: 02 Mar 2016 *

Full-text access for editors Access for subscribers Purchase this article Comment on this article