An Algorithm for Ranking Web Pages Based on Links and Ant Colony Algorithm

Abstract

With the exponential web growth, techniques of recommender systems and pages ranking algorithms have gained importance over time. Web mining which is considered as a subset of data mining, is being emphasized in three categories: web mining based on application, web mining based on content and web mining based on structure. Pages ranking algorithms operates mainly based on structure-based web mining. In the current study it has been tried to maximize accuracy of pages ranking with combining function and structure-based techniques. Improvement of Page Rank algorithm is performed using user profiles and important attributes to page ranking algorithm (number of inbound and outbound links to pages). Due to the problem broadness, the use of meta-heuristic algorithms (such as ant colony algorithm) has been highlighted in the current study and the fitness function is set so that the increase of iterations will increase the accuracy of PageRank algorithms. The results of the study explain accuracy of the proposed method compared to other methods.