Email-based Spam Detection using Long Short-Term Memory

Spam emails are a major problem in today's digital world because they fill users' inboxes with unsolicited and often dangerous content. Robust techniques for correctly detecting and removing spam messages are needed to solve this problem. This paper suggests a strategy to categorize emails as spam or legitimate that is based on supervised learning and unstructured text processing. The goal of the project is to preprocess email data and extract pertinent features for classification by utilizing techniques including tokenization, stop word removal, stemming, and feature extraction. Our goal is to create an efficient spam classifier by analyzing and fine-tuning with Naïve bayes, Long Short-Term Memory (LSTM). Performance measures including precision and recall will be applied for computing the efficiency of the classifier. Additionally, this presented work improves the efficiency of LSTM-based learning methodologies for email spam detection.