Authors: T.V. Vijay Kumar; Kumar Dilip
Addresses: School of Computer and Systems Sciences, Jawaharlal Nehru University, New Delhi-110067, India ' School of Computer and Systems Sciences, Jawaharlal Nehru University, New Delhi-110067, India
Abstract: It has been observed that organisations in order to be competitive are trying to maintain long-term relationships with their online customers. These organisations are mining information in log files or on-line forms with the aim of creating websites that can adapt to changing online buying patterns of customers. In addition to improving the browsing experience of the user, this would also lead to an increase in sales of the product. This problem of arranging web objects in an optimal manner, on a website, with respect to download time, visualisation and potential sales of the product, has been addressed as a tri-objective optimisation problem with the three objectives being minimising the download or waiting time, maximising the product visualisation and maximising the potential sales of the product. The proposed NSGA-II-based multi-criteria website optimisation algorithm is able to generate multiple web object sequences, with each providing optimal trade-offs between the three objectives resulting in the effective and efficient catering of the varied class of users.
Keywords: web usage mining; website adaptation; multi-objective optimisation; non-dominated sorting genetic algorithms; NSGA-II; website optimisation; download time; product visualisation; potential sales; waiting time; web object sequences.
International Journal of Business Information Systems, 2016 Vol.21 No.4, pp.418 - 438
Received: 26 Jul 2014
Accepted: 11 Oct 2014
Published online: 09 Mar 2016 *