An approximation approach to discovering web services for uncertain client's qos preference

It is paramount to provide seamless and ubiquitous access to rich contents available online to interested users via a wide range of devices with varied characteristics. However, mobile devices accessing these rich contents are constrained by different capabilities e.g., display size, thus resulting poor browsing experiences e.g., unorganized layout. Recently, a service-oriented content adaptation (SOCA) scheme has emerged to address this content-device mismatch problem. In this scheme, content adaptation functions are provided as services by multiple providers. This elevates service discovery as an important problem. A QoS-based service discovery approach has been proposed and widely used to matchmaking the client QoS preference with the service advertised QoS. Most of these solutions assume that the client's QoS is known a priori. However, these approaches suffer from unknown or partially specified client QoS. In this paper, we propose an approximation approach to deal with QoS uncertainty. Our solution considers the statistical approach to discover the suitable content adaptation services. The performance analysis verifies that our approach performs reasonably well.