Causal analysis of attacks against honeypots based on properties of countries

摘要

This study studies the influence of country attributes on the number of secure shell attacks originating from it detected by the author's honeynet. Four statistical models are described, based on three sources of data from various countries. The studied attributes of the countries can be broadly divided into demographic, technological, and economic, with each source providing a slightly different set of attributes. Statistical methods such as partial least-squares path modelling are used, clustering countries by their assessed similarity. The population size has the greatest effect on the number of attacks, as expected, though it has to be noted that developing countries did not provide relevant data to the sources used and thus were not included. The following influential attributes were technical such as the access to information and communication technologies (ICT), and the use of ICT, with the economic influence being notable only in rather small countries. The Netherlands was an interesting anomaly, being clustered alongside large countries, even though its country attributes were very much like those of its neighbours.