Detecting phishing e-mails using text and data mining

• Published in: 2012 IEEE International Conference on Computational Intelligence and Computing Research pp. 1 - 6
• Main Authors: Pandey, M., Ravi, V.
• Format: Conference Proceeding
• Language: English
• Published: IEEE 01.12.2012
• Subjects: Classification; Decision Tree; Genetic Programming; Group Method Of Data Handling; Logistic regression; Multilayer Perceptron; Phishing webpage; Probabilistic Neural Network; Support Vector Machine; Text mining
• ISBN: 1467313424; 9781467313421
• DOI: 10.1109/ICCIC.2012.6510259

This paper presents text and data mining in tandem to detect the phishing email. The study employs Multilayer Perceptron (MLP), Decision Trees (DT), Support Vector Machine (SVM), Group Method of Data Handling (GMDH), Probabilistic Neural Net (PNN), Genetic Programming (GP) and Logistic Regression (LR) for classification. A dataset of 2500 phishing and non phishing emails is analyzed after extracting 23 keywords from the email bodies using text mining from the original dataset. Further, we selected 12 most important features using t-statistic based feature selection. Here, we did not find statistically significant difference in sensitivity as indicated by t-test at 1% level of significance, both with and without feature selection across all techniques except PNN. Since, the GP and DT are not statistically significantly different either with or without feature selection at 1% level of significance, DT should be preferred because it yields `if-then' rules, thereby increasing the comprehensibility of the system.