Title: Learning from socio-economic characteristics of IP geo-locations for cybercrime prediction

Authors: Keivan Kianmehr; Negar Koochakzadeh

Addresses: Department of Electrical and Computer Engineering, Western University, London, Ontario, Canada. ' Department of Computer Science, University of Calgary, Calgary, Alberta, Canada

Abstract: Cybercrime detection solutions have recently received increased attention. Predicting the cybercrime potentiality of a request received by a server can reduce the risk of cybercrime. In this paper, we present an alternative solution to the current intrusion detection systems in that the socio-economic characteristics of IP geo-locations of a request are used to predict its crime potentiality. The IP address of a request is used to exploit its socio-economic characteristics. Using the IP address of a request, the physical location, from where the request has been sent, is identified. Socio-economic attributes of people living in that area are collected. These characteristics can specify the seriousness of a cybercrime associated with a request. Classification algorithms can be used to build a prediction model. We have conducted a case study in which we built a prediction model using a set of socio-economic attributes. Our results show the applicability of the proposed model.

Keywords: cyber security; cybercrime prediction; learning; socio-economic characteristics; IP geolocation; crime detection; intrusion detection; IP address; physical location; classification algorithms; prediction modelling; cyber attacks; server security; data mining.

DOI: 10.1504/IJBIDM.2012.048726

International Journal of Business Intelligence and Data Mining, 2012 Vol.7 No.1/2, pp.21 - 39

Published online: 12 Nov 2014 *

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