Classification of Fake News Using Enhanced Capsule Neural Network

Recognition of Fake News (RFN) was based on vital domain within Deep Learning (DL) along with Machine Learning (ML), applied in various contexts, such as journalism, social media platforms, online news portals, and fact-checking organizations. Providing incorrect information misleads the public, causes objections, and generates concerns for both individuals and the authorities. It was necessary to check the validity within the information during the beginning prior releasing it to everyone. Deep learning, including Capsule Neural Network. The future application of Capsule Neural Network in news verification systems holds great promise in combating the spread of misinformation, thereby fostering a more informed and resilient society. The purpose of the analysis was to apply Capsule Neural Network in the identification of fake news. Capsule Neural Network is a custom Keras layer representing a Capsule Network. It takes the number of capsules, capsule dimensions, and routing iterations as parameters. Inside the Capsule Network layer, it includes a Capsule Layer that handles the primary capsule layer's operations. The fake news detection model includes 6 layers: Input Layer, Embedding Layer, Convolution Layer, Flatten Layer, Capsule Layer and Output Layer.