9th International Conference on ICT for Sustainable Development

Authors - Linda Varghese, Rajesh R Pai, Nandini Kumari, Savitha G, Girisha S
Abstract - The unsolicited and misleading email material is sent in bulk to many recipients, sometimes known as spam or junk email. In the recent time, the increasing volume of such emails poses challenges to the electronic communication. This study therefore tries to build a reliable and accurate method for identifying and preventing spam emails, for improving user experience and information security. The dataset "Spam email classification" extracted from the Kaggle website is used in this study to detect and categorise email spam. It analyses the text of the email using natural language processing and applies machine learning techniques to original unbalanced and resampled balanced datasets. The results indicate that the Random Forest model performs most effectively with an F1-Score of 98% and an accuracy of 93%, respectively.