In this work we have extended and implemented an ontology based approach for email classification based on user characteristics proposed by Kim et al.(2007). The approach focuses on finding relationships between user interests and their responses to emails. Rules and Ontology are created using the data and metadata of user characteristics, their preferences and responses to emails. Rules and ontology are then used to predict the response of a user to a new email. In Kim et al. (2007) approach, labels to emails were provided manually by a human expert. We have endeavored to remove the human intervention by developing an Automated Email Categorizer to provide label to an email based on its contents. We have also proposed a new term weighing method for emails to incorporate prominence of subject terms. Finally, we have integrated and tested the Ontology Based Classifier in conjunction with Email Categorizer where the former effectively uses the label provided by latter to classify an email based on user preferences.