Web Page Recommendation System Using Laplace Correction Dependent Probability and Chronological Dragonfly-Based Clustering

Abstract

The web page recommenders predict and recommend the web pages to the users based on the behavior of their search history. The web page recommender system analyzes the semantics of the navigation by the user and predicts the related web pages for the user. Various recommender systems have been developed in the literature for the web page recommendation. In the first work, a web page recommendation system was developed using weighted sequential pattern mining and Wu and Li Index Fuzzy clustering (WLI-FC) algorithm. In this work, the Chronological based Dragonfly Algorithm (Chronological-DA) is proposed for recommending the webpage to the users. The proposed Chronological-DA algorithm includes the concept of the chronological for recommending the webpage based on the history of pages visited by the users. Also, the proposed recommendation system uses the concept of Laplacian correction for defining the recommendation probability. Simulation of the proposed webpage recommendation system with the chronological-DA uses the standard CTI and the MSNBC database for the experimentation, and the experimental results prove that the proposed scheme has better values of 1, 0.964, and 0.973 for precision, recall, and F-measure respectively.