An Automated Learning Of Unsolicited Mail Detection Using Bilstm And Gfgsc Classifier With Goa To Proliferate Accuracy

Download or read book An Automated Learning of Unsolicited Mail Detection Using BiLSTM And GFGSC Classifier With GOA To Proliferate Accuracy written by N. A. S. Vinoth and published by Mohammed Abdul Sattar. This book was released on 2024-05-27 with total page 0 pages. Available in PDF, EPUB and Kindle. Book excerpt: Email spam, also known as unsolicited bulk email (UBE), is the term used to describe the mass distribution of unwanted or irrelevant messages to a large number of recipients via email. These messages are typically sent by commercial entities or individuals with the intention of promoting a product or service, scamming recipients, or spreading malware. Spam emails often contain misleading subject lines, deceptive content, and fake sender information. They can also be used for phishing attacks, in which the sender attempts to trick recipients into providing personal information or clicking on malicious links. Email spam is a widespread problem that affects individuals, businesses, and organizations of all sizes [1]. It can clog up inboxes, waste time and resources, and pose a security risk to users. As a result, many email providers have implemented various spam filtering techniques to automatically detect and block spam messages before they reach the recipient's inbox. Despite these efforts, spammers continue to find new ways to evade filters and send unwanted messages, making email spam an ongoing challenge for internet users [2]. Deep learning is a powerful technique that can be used to build robust email spam detection systems. The deep learning model involved in automated learning from the pattern those are complex with the raw data those are suitable for the spam detection. An automated system to detect unsolicited mail (spam) using advanced machine learning techniques, leverages a Bidirectional Long Short-Term Memory (BiLSTM) network to analyze the sequential nature of email content, improving the capture of contextual information from both past and future data points. To further enhance classification accuracy, the system integrates a Gradient-Frequency-based Scalable Classifier (GFGSC), which refines the decision boundaries based on the gradient and frequency of features. Additionally, the system employs the Grasshopper Optimization Algorithm (GOA) to optimize the hyperparameters of the BiLSTM and GFGSC models.