A client perceived QoS model for web services selection

Nowadays, a number of web services available on the Internet have been increasing rapidly. Similarly functional services are offered by different service providers. Quality of service (QoS) has becomes an important factor to determine the most suitable service provider. Many existing QoS models use uncertain QoS attributes which vary over time and depend on service environment. The service chosen by those models may not provide a QoS as expected. In this paper, we present a QoS model using the uncertain attributes: response time, availability and reliability. In order to accurately selecting a suitable service from these uncertain attributes, our model uses a non-parametric test technique which tests historical QoS information that measured from client side.