Email Spam Detection Using BiLSTM with Large Scale Hybrid Datasets

Email has become one of the most widely used communication platforms in modern digital society. However, the rapid growth of email communication has also led to a significant increase in unsolicited and malicious emails commonly referred to as spam. Spam emails are not only a source of inconvenience for users but also a major channel for cyber threats, including phishing attacks, malware distribution, financial fraud, and identity theft. Traditional spam filtering techniques, such as rule-based filtering and conventional machine learning algorithms, often rely on keyword matching or statistical word frequency models. Although these approaches have shown moderate success, they fail to effectively capture the contextual meaning and sequential structure of natural language, allowing sophisticated spam messages to bypass traditional filters. This research proposes an advanced deep learning-based email spam detection system using a Bidirectional Long Short-Term Memory (BiLSTM) architecture to overcome the limitations of traditional approaches. The primary objective of this study is to develop a context-aware spam classification model that accurately distinguishes spam from legitimate.