Offensive Message Spam Detection in Social Media Using Long Short-Term Memory

In this world, online communication has become an essential part in our daily lives. Almost everyone uses social media platforms to communicate with the whole world. But this also leads to a major problem: cyberbullying. This paper proposes a system to detect users who spam offensive messages repeatedly and ban them using natural language processing and machine learning. The system would be beneficial to service providers, social media platforms, and individuals, allowing us to strive for a world of harmony. The amalgamation of NLP, Big Data, and ML approaches presents a robust and efficient solution for offensive message spam detection. The findings of this project will shed light on the potential of these technologies to curb the proliferation of offensive content, thereby fostering a healthier and more respectful online communication ecosystem. This paper delves on offensive spam detection using the LSTM model, and the obtained MSE of this model is 12.58.