Modeling Web maintenance centers through queue models

Abstract

The Internet and the World Wide Web's pervasiveness are changing the landscape of several different areas, ranging from information gathering/managing and commerce to software development, maintenance and evolution. Traditional telephone-centric services, such as ordering of goods, maintenance/repair intervention requests and bug/defect reporting, are moving towards Web-centric solutions. This paper proposes the adoption of queuing theory to support the design, staffing, management and assessment of Web-centric service centers. Data from a mailing list archiving a mixture of corrective maintenance and information requests were used to mimic a service center. Queuing theory was adopted to model the relation between the number of servers and the performance level. Empirical evidence revealed that, by adding an express lane and a dispatcher service time, the variability is greatly reduced and more complex business rules may be implemented. Moreover, express-lane customers experience a reduction of service time, even in the presence of a significant percentage of requests erroneously routed by the dispatcher.