Building dynamic e-business systems architecture

Abstract

Web personalization is the process of customizing a web site as per user‘s preferences by using the data mining, taking advantage of the knowledge acquired from the analysis of the user‘s navigational behavior. With personalization, when a particular person visits a Website, the recommendation can be designed specifically for that person. The Dynamic E-Business website can be designed using web personalization. The proposed E-Business Web personalization model consists of preprocessing, Knowledge Discovery using data mining and recommendation engine. The proposed preprocessing algorithm consists of three subroutines, namely Cleaning non-required server log, Identification of User from IP address and Log, Creating session for every user and preparing Transaction file. The knowledge discover module consists of Aprori Algorithm for TRP association rule, Sequence Pattern Algorithm for Mining log and Access Log and ID3 Algorithm for Classification of the registered Users of the website. Depending on the Sequence pattern, association rule and user class the recommendation module recommends the desired web page link to the user