一种综合考虑主客观权重的Web服务QoS度量算法∗

摘要

The existing methods for measuring Web service QoS (quality of service) are not accurate as they cannot precisely quantify the ambiguity of user preference while neglecting the characteristics of QoS data set. This paper presents a QoS measuring algorithm which employs subjective and objective weight. The new approach can automatically adapt to user preference with a subjective weight calculation method, and can evaluate service performance accurately with an objective weight calculation method. The algorithm takes a comprehensive consideration of both subjective and objective aspects to measure Web service QoS, therefore the measure results can conform to user preference and reflect the overall service performance accurately. The theoretical analysis and experimental results based on the QWS real data set show that the proposed algorithm can measure Web service QoS accurately.%已有的Web服务QoS(quality of service)度量方法由于无法对用户偏好的模糊性予以准确量化，以及对候选服务 QoS 属性数据分布特征的忽视，导致其度量结果不准确。为此，提出了一种综合考虑主客观权重的 Web 服务QoS 度量算法。该算法利用自适应用户偏好的主观权重计算方法和服务潜能保障的客观权重计算方法，从主观和客观两个角度进行QoS度量，以保障度量结果在符合用户偏好的基础上能够准确地反映服务的整体性能。理论分析和基于QWS真实数据集的实验结果表明，所提出的方法能够准确地获得Web服务QoS的度量结果。