In the rapidly evolving landscape of digital technology, the Internet has become an indispensable tool for communication and information sharing. Platforms like Quora play a pivotal role in this digital age, enabling users to ask and answer questions on a wide range of topics. However, ensuring the quality and authenticity of content on Quora remains a challenging task. This research paper delves into the critical task of classifying messages on Quora, distinguishing between useful and irrelevant content, and assessing the authenticity of user-generated questions. The current system employed on Quora falls short of achieving optimal accuracy in these tasks, prompting the development of an enhanced system leveraging machine learning techniques. This study utilizes Naïve Bayes, logistic regression, and support vector machine algorithms to create a robust model. Remarkably, the proposed system achieves an impressive accuracy rate of up to 99.87%, effectively addressing the limitations of the existing system. This research highlights the potential of machine learning in improving content quality and authenticity assessment on question-and-answer platforms contributing to a more reliable and informative online environment.