Web Personalization With Web Usage Mining Technics and Association Rules

Abstract

As amount of information and web development increase considerably, some technics and methods are required to allow efficient access to data and information extraction from them. Extracting useful pattern from worldwide networks that are referred to as web mining is considered as one of the main applications of data mining. The key challenges of web users are exploring websites for finding the relevant information by taking minimum time in an efficient manner. Discovering the hidden knowledge in the manner of interaction in the web is considered as one of the most important technics in web utilization mining. Information overload is one of the main problems in current web and for tackling this problem the web personalization systems are presented that adapts the content and services of a website with user's interests and browsing behavior. Today website personalization is turned into a popular event for web users and it plays a leading role in speed of access and providing users' desirable information. The objective of current article is extracting index based on users' behavior and web personalization using web mining technics based on utilization and association rules. In proposed methods the weighting criteria showing the extent of interest of users to the pages are expressed and a method is presented based on combination of association rules and clustering by perceptron neural network for web personalization. The proposed method simulation results suggest the improvement of precision and coverage criteria with respect to other compared methods.