Machine Learning Email Prediction System (MLEPS)

Abstract

Nowadays, email has become one of the most critical personal and business applications and email users would experience serious consequences if email messages could not be available or experience high volume of messages which lead to congestions, overloads and limited storage space coupled with unstructured messages in mail boxes. A few years ago, the means of communication are via letters by post, telegraph, fax, couriers to mention a few but now the focus has changed to a faster means of obtaining quick responses and faster ways of communication-emails. We propose a new framework to help organise and prioritize email better; Machine learning email prediction system (MLEPS). The goal is to organise emails better in mail boxes, prioritise emails based on the focus of the email content. The intelligent email prediction system helps to improve email users’ performances, saves time, very effective and efficient tool and is cost effective for businesses and for personal use. The system is evaluated against a corpus of human-judged predictions, reaching satisfactory level of performance.