ML-EC2: An Algorithm for Multi-Label Email Classification Using Clustering

Literature Review

Managing huge amount of emails received from users is a very challenging problem which needs to be solved in an effective and efficient way. Various researches have been done in the field of email mining. Some of the surveys done are as follows.

Park & An (2010) proposed an Email multicategory classification approach using semantic features and a dynamic category hierarchy reconstruction method in which the user reorganizes all e-mail messages into categories. Guan & Yuan (2013) reviews the existing work on mislabeled data detection techniques for pattern classification and classifies them into three types: Local learning-based, ensemble learning-based and single learning-based methods. The author Armentano & Amandi (2014) presented an approach to label the incoming emails based on user preference; a set of experiments using Google’s webmail system, Gmail is performed to obtain a good rate of acceptance of the agent interactions. Alsmadi & Alhami (2015) introduced an algorithm for performing clustering and classification of email text corpus. They have proposed a model for classification of emails based on subject and folder using N-grams. Islam et al. (2009) proposed a new technique of e-mail classification based on the analysis of grey list (GL), which uses multi-classifier classification ensembles of statistical learning algorithms.