Score based Support Vector Machine for Spam Mail Detection

Nowadays, people are well aware of the essential role that social media plays in modern communication. Email is one of the most reliable and secure social media platforms used for making online contacts and exchanging data or messages over the internet. Free email and messages provided by strangers are relied on by many people nowadays. Because anyone may send an email or a message, spammers have a fantastic opportunity to write spam messages that cater to our varied interests. Mass emailing, sometimes known as “email spam,” is the practice of sending numerous messages at once using electronic mail. When emails are forwarded to unauthorized servers, spam is the usual result. Spam slows down internet connection, steals sensitive information like passwords and credit card details, and tampers with search results on computers. It's not simple to track down spammers and their content. Today, spam emails are mostly used to promote products and services, as this is where the money is. The proposed model allows us to categorize messages as spam or ham and determine what percentage of the dataset is comprised of spam or ham. In this proposed work, the machine learning method such as Support Vector Machine (SVM) are examined and how it is used to classify emails as spam or ham using proposed dataset. The accuracy of suggested model for determining the content of an email is significantly higher than that of earlier models.