Social media's proliferation of fake news had serious consequences in the real world, raising concerns among net users worldwide in recent years. Detection mechanisms for deception have also piqued the interest of researchers around the world. There has been a recent uptick in public and academic interest in spotting fake news, thanks to the proliferation of misinformation disseminated through online media such as social media feeds, electronic newspapers and news blogs. The most recent information, however, is dubious and frequently misleads other users on an online social network. It is challenging to identify fake news based on shared material since fake news is deliberately spread to lead readers astray and convince them to believe misleading information. As a result, some new data to the user's profile, such as their participation in a particular decision-making process must be added. The need for fake news identification is highlighted by the difficulty in quickly recognizing these contents due to the spread of information and its process. In this paper, the fake news detection model is developed by considering optimal features such as Article Performance Features (APF) and news features using machine learning and deep learning models. The experimental study of the proposed model shows that the intended strategy for detecting fake news has greater accuracy than the current state-of-the-art methods.

Fake News, Optimal Features, Machine Learning, Deep Learning, LSTM