International Journal of Scientific Research and Engineering Development

Spam message classification has become a crucial task in ensuring secure and reliable digital communication, especially with the exponential growth of mobile messaging and email services. Traditional machine learning models such as K-Nearest Neighbour (KNN), Support Vector Machine (SVM), and Random Forest have been widely applied; however, these models often rely heavily on feature engineering and may not efficiently capture sequential dependencies in text. To overcome these limitations, this study employs a Long Short-Term Memory (LSTM) based deep learning approach for spam detection. LSTM networks are well-suited for natural language processing tasks due to their ability to retain long-term dependencies and context within sequential data. By leveraging word embeddings and an LSTM architecture, the system automatically learns meaningful text representations and classifies messages as spam or ham with high accuracy. This method reduces reliance on handcrafted features and demonstrates superior adaptability to diverse linguistic patterns.