This web prediction model based on the structure of the Recurrent Neural Network(RNN) is used for performing web page predicting during the process of users browsing the web. The most important purpose of this approach is to reduce latency when users browse the web by predicting the user's next purpose and preloading the webpage ahead of time.
Prediction process is consist of the following 5 phases: preprocessing, encoding, clustering, predicion, bundling. Files related to each phase are in directories corresponding to the name of each.