EnronEmailCategorization

Emails are important and pervasive in people lives as a large section of the society exchange digital messages as a means for communication. Therefore, it is very important to keep this chunk organized and categorized as large volumes of different types of emails result in a cluttered mailbox and in this messy outlook, some important emails may go unnoticed. I've performed email categorization based on the body of the email using real world email dataset provided by Enron Corporation which consisted of over 500000 emails from over 150 people’s inboxes. I've performed experiments and on a chunk of this enron dataset which was labeled for its categories and since the volume of the dataset was so large, it could be easily divided into training, testing and validation sets which were modeled using Naive Bayes, Linear Support Vector and K-nearest neighbors classifiers to provide baseline results using text extractors namely TF-IDF Vectorizer and Count Vectorizer.