Bot detection plays a crucial role in maintaining the integrity and trustworthiness of online plat-forms, especially in the context of social media. This paper presents a deep learning-based methodology for bot detection on Twitter. We collected a dataset of 10,000 tweets and conducted rigorous data preprocessing to ensure data quality. Various deep learning architectures, including LSTM, CNN, and GNN, were explored and adapted for bot detection. Relevant linguistic, temporal, user-related, and network-related features were extracted to capture different aspects of bot behavior. Through extensive training and evaluation, our methodology achieved an accuracy of 0.90, precision of 0.91, recall of 0.89, and F1 score of 0.90, demonstrating its effectiveness in accurately identifying bot accounts. Comparative analysis with other studies using the same dataset revealed superior accuracy, precision, recall, and F1 score performance. The implications of this research are significant in domains such as social media, cybersecurity, and online platforms. By effectively identifying and mitigating the presence of bots, we can enhance the authenticity and trustworthiness of online conversations. However, challenges and limitations must be addressed when adapting the methodology to evolving bot behaviors and generalizing it to different datasets and platforms. This paper contributes to bot detection by presenting a comprehensive method that combines deep learning architectures, feature extraction techniques, and rigorous evaluation, opening avenues for further advancements in bot detection research.