A web service has many different QoS attributes. Different customers have different preferences on different QoS attributes of a web service. For getting a service which can best fit the customer's desire, this paper gives a quantitative selection algorithm based on the weighted space distance in the n dimensions. Using the preferences, the algorithm calculates the first-order weighted space distance of every service, and selects the best service named "a"; If the relative dispersion is greater than or equal to the threshold value, it returns service "a" to the customer. Or else, the algorithm calculates the second-order weighted space distance of every service, and selects the best service named "b"; if the relative dispersion is greater than or equal to the threshold value, it returns service "b" to the customer. Otherwise the service returns to the customer is the one that has the large relative dispersion. This algorithm can select the best one from web services with the same performance, which sufficiently indicates the preferences, and thus has its obvious advantage.%一个Web Service服务有很多的QoS属性,不同的用户对不同的QoS有不同的偏好要求,为从众多功能相同的Web Service中选择用户最期望的服务,本文给出了基于n维空间的权重距离的算法.根据用户给定的偏好,计算每个Web Service服务的一阶权重距离,其最大值对应的服务a为最优；如果服务a和次优服务的相对差大于等于给定阈值,则向用户返回该服务；否则计算二阶权重距离对应的最优服务b,如果服务b和次优服务的相对差大于等于给定阈值,就向用户返回服务b；否则,比较服务a和服务b的相对差,将相对差较大的服务返回给用户.该算法能充分体现用户对不同QoS属性的偏好,因而具有明显的优势.
展开▼
