Fraud and terrorism have a close connect in terms of the processes that enables and promote them. In the era of Internet, its various services that include Web, e-mail, social networks, blogs, instant messaging, chats, etc. are used in terrorism not
only for communication but also for i) creation of ideology, ii) resource gathering, iii) recruitment, indoctrination and training, iv) creation of terror network, and v) information gathering. A major challenge for law enforcement and intelligence agencies is efficient and accurate gathering of relevant and growing volume of crime data. This paper reports on use of established Naive Bayesian filter for classification of threat e-mails. Efficiency in filtering threat e-mail by use of three different Naive Bayesian filter approaches i.e. single keywords, weighted multiple keywords and weighted multiple keywords with keyword context matching are evaluated on a threat e-mail corpus created by extracting data from sources that are very close to terrorism.
In this paper we discuss the techniques involved in the design of the famous statistical spam filters that include Naive Bayes, Term Frequency-Inverse Document Frequency, K-Nearest Neighbor, Support Vector Machine, and Bayes Additive Regression Tree.
We compare these techniques with each other in terms of accuracy, recall, precision, etc. Further, we discuss the effectiveness and limitations of statistical filters in filtering out various types of spam from legitimate e-mails.
