Recurrent neural network-based user identity misappropriation prevention from public domains and connections

Abstract

Web browsing security is realized through use of a recurrent neural network (RNN) capable of creating a plurality of layers comprising web browser session wrappers, which overlay the actual user session. In this regard, the RNN creates hidden layer(s) and a final layer overlaying the last hidden layer which is exposed to the website and or domain. The RNN implements homomorphic encryption to systematically encrypt the session data and/or user data associated with the actual user session and stored or accessible to the various session wrappers. Once unauthorized use/tampering is detecting in the final layer, the RNN can reconfigure/add to the existing hidden layers and/or the homomorphic encryption can re-encrypt the session/user data is each layer/session wrapper.