Title: Spam detection using hybrid model of rough set and decorate ensemble
Authors: Sanjiban Sekhar Roy; V. Madhu Viswanatham; P. Venkata Krishna
Addresses: SCOPE, VIT University, SJT-116-A29, Vellore, Tamil Nadu, 632014, India ' SCOPE, VIT University, SJT-411-A25, Vellore, Tamil Nadu, 632014, India ' Sri Padmavati Mahila Visvavidyalayam, Padmavathi Nagar, Near West Railway Station, Tirupati, Chittoor (D.t), Andhra Pradesh, India
Abstract: The amount of junk emails popularly known as spam emails is ever increasing. It has created inconvenience for email users. Moreover, spam emails are causing huge monetary loss for the organisations. This situation has compelled researchers to build effective spam filters. This paper proposes a hybrid model of rough set and decorate ensemble for the detection of spam emails. Experimental outcome demonstrates that the framework using 'rough set hybridised with decorate ensemble' attains good precision and recall. The outcome of the proposed method surpasses the performances of the existing ensembles. Selecting the optimal value of the parameters is a key criterion and this was achieved by performing ten-fold cross-validations.
Keywords: junk emails; spam detection; filters; rough sets; decorate ensemble; cross validation.
DOI: 10.1504/IJCSYSE.2016.079000
International Journal of Computational Systems Engineering, 2016 Vol.2 No.3, pp.139 - 147
Received: 13 Jul 2015
Accepted: 26 Feb 2016
Published online: 08 Sep 2016 *