IRMA-International.org: Creator of Knowledge
Information Resources Management Association
Advancing the Concepts & Practices of Information Resources Management in Modern Organizations

Automating Web Personalization with a Self-Organizing Neural Network

Automating Web Personalization with a Self-Organizing Neural Network
View Free PDF
Author(s): Victor Perotti (Rochester Institute of Technology, USA)
Copyright: 2003
Pages: 2
Source title: Information Technology & Organizations: Trends, Issues, Challenges & Solutions
Source Editor(s): Mehdi Khosrow-Pour, D.B.A. (Information Resources Management Association, USA)
DOI: 10.4018/978-1-59140-066-0.ch342
ISBN13: 9781616921248
EISBN13: 9781466665330

Abstract

The Internet’s continuing transformation of business has created new and unique demands for information management. While these demands are multifaceted, perhaps none are more important than pursuing an understanding of website users, and leveraging this information to create site structure and content that is appropriate for users. Indeed, personalization and customization of web services are now commonplace features on many e-Business sites. Certainly, many sites use “cookies” or other technologies to track each individual as they come to the site. Consumers, however, are increasingly wary when their personal information is requested. A less invasive approach is to look at the aggregate behavior of all users, and to try to identify trends therein. Once these trends are identified, a user can be classified as a member of a particular group, and customized web content can be delivered. Recommendation engines that identify user trends, and deliver user-appropriate content are an active field of research today. For example, Perkowitz & Etzioni (1997) challenged the Artificial Intelligence community to develop truly adaptive web sites that respond to the behavior of their users. Their more recent work (1998) introduces conceptual analysis as part of a “cluster mining” process which identifies groups of users that have common usage profiles. Similarly, Mobasher, Cooley and Srivastava. (2000) have developed an entire process for clustering web users, and have shown its performance relative to other algorithms. While the recovery of usage profiles by these two groups is fairly advanced, the degree of adaptation offered is fairly small.

Body Bottom