Manipulation of Email Data Using Machine Learning and Data Visualization

Abstract

E-mail has become one of the essential economic for all forms of communication in today's life. The rise in the users of email has drastically increased the data set of the email available on the one tap over the internet. In this paper we will propose an algorithm based on machine learning which will classify the email based on its subject. We have used several machines learning algorithms classifier Such as SVM classifier, neural network classifier. however people mostly prefer email to be as a communication for business and other personal purposes. Application of the emails has been used everywhere in education, corporate, business and so on. With the Rise of the data set of the email it’s generate a Corpus with itself which can be used as a different categorization through which we will classify the email based on its subject matter. The rise in the number of the data sets of the email it brings some more features along with it through which we can extract some features with it and we can implement opinion mining and sentiment analysis and thereby we can extract spam ham detections of the email data out of it. We have used supervised machine learning algorithm for the implementation of the data sets that we have used and converted the unlabeled and unstructured data set of email into the labeled and structured datasets and then we have extracted the features from it. Moreover various public data sets, feature sets, classification techniques, performance measures are examined and use in each in identified application area. In this paper we have used several datasets of email for the subject based classification and we have also proposed algorithm for spam detection for this method we have employed several machine learning algorithm.