E-mail Classification Using Deep Networks Ahmad

Email has become an essential communication tool in modern life, creating the need to manage the huge information generated. Email classification is a desirable feature in an email client to manage the email messages and categorize them into semantic groups. Statistical artificial intelligence and machine learning is a typical approach to solve such problem, driven by the success of such methods in other areas of knowledge management. Many classification methods exist, of which some have been already applied to email classification task, like Naïve Bayes Classifiers and SVM. Deep architectures of pattern classifiers represent a wide category of classifiers. Recently, Deep Belief Networks have demonstrated good performance in literature, driven by the fast learning algorithm using Restricted Boltzmann Machine model by Hinton et al, and the improvement in computing power which enables learning deep neural networks in reasonable time. Many datasets and corpus exist for email classification, with the most famous one is Enron dataset, made public by FERC, annotated and processed by many entities like SRI and MIT. In this paper, a machine learning approach using Deep Belief Networks is applied to email classification task, using the Enron dataset to train and test the proposed model.