Enhancing Web Page Recommendation Accuracy via Session Clustering and Backpropagation Neural Networks on CTI Dataset

Abstract

Recently, several studies have been conducted to improve the quality of web page recommendation systems (WPRs) using Web Usage Mining (WUM). Web page RS is an intelligent technique that generates page recommendations through systematic processing of anonymous web user navigation data. These are then categorized and presented to the active user, in the form of links, which may be of interest to the user without their explicit request. These systems are evaluated by measuring the accuracy of predicting future user decisions. Accuracy is a measure of the effectiveness of the system based on the proposed solution, which reflects user satisfaction. However, these current solutions still do not fully meet the user's desires. In this work, we present a study of a web usage data mining system to improve the prediction accuracy of a web page recommendation system. To achieve this, a back-propagation algorithm based on an artificial neural network was used to improve the quality of the detected user session set. The study relies on conducting several experiments on the CTI dataset, and the results showed an improvement in the quality of Web page recommendation.