Comparison of Decision Tree Algorithms for Spam E-mail Filtering

Spam e-mails are quickly becoming a serious problem that faces the internet community, threatening the efficiency of users. Spam e-mails violates private information, which is costly and unsolicited communication. The intrusion of spam e-mails pursues the users and waste the network resources. Spam filtering Off er a wide range of methods aimed to detect spam emails, these are implemented in many ways and at several levels. New technologies, are improving filter accuracy. Several filtering methods have been implemented so far, mainly based on different Machine Learning algorithms to filter out spam e-mails. However, some of those methods gives poor accuracy and some of them are costly regarding to computational complexity. In this paper, we propose an approach for spam e-mail filtering based on decision tree algorithms which are simple and gives better accuracy. From experimental results, the proposed random forest classifier outperforms other decision tree methods for publicly available datasets.