Website user content navigation behavior modeling using time series neural networks.

User navigation behavior modeling has attracted much research interest on the web. User Navigation Behavior modeling deals with creating a model to predict future user click based on previous user clicks. Previous works focus on predicting the next page without much attention to the user-selected content. In contrast to existing works, this paper focuses on the user selected contents and the page content layout for creating the user navigation behavior. In this paper, for the first time in the literature, a novel model for user navigation prediction has been proposed using the time series neural network. The performance of the proposed prediction model was examined by two variations of time series neural networks including nonlinear autoregressive (NAR) and nonlinear autoregressive with exogenous input (NARX). Providing external information for the NARX, page content layout is used. The experimental results show that the NAR and NARX neural networks are more efficient than other methods for click prediction, and using the page content layout in the NARX improves the accuracy of predictions. [ABSTRACT FROM AUTHOR]