领域自适应的Web服务评价模型

Web服务质量的评价是指导Web服务的选取与组合的主要手段，而目前的Web服务评价模型采用与服务领域无关的评价因子来评价服务，不能全面衡量服务使用者对Web服务质量的要求．该文用基于语义的方法描述Web服务评价模型，支持动态定制不同领域的服务评价因子，同时利用Web服务在使用过程中产生的知识以及领域专家的领域知识交互计算评价因子的权重分布．实验证明，该模型的评价结果优于其它评价模型的评价结果，并且采用交互算法的准确率比仅采用机器学习算法或仅使用先验知识的准确率提高10％以上．该模型可应用于Web服务管理系统、Web服务选取和Web服务合成引擎等Web服务应用系统中．

Domain -Adaptive Service Evaluation Model

The accuracy of service quality evaluation is the key to selecting Web service from many available ones which provide overlapping or identical functionality, but current Web service evaluation models evaluate it only by common service attributes, which can not meet the requirement from Web service requestors. This paper presents a new Web service evaluation model described by Semantic language, which supports to customize evaluation attributes dynamically. It computes the weight of each evaluation attribute by utilizing both the machine learning algorithm on the knowledge generated from Web service runtime environment and active algorithm on transcendent knowledge from domain experts. The evaluation of relative experiments shows this model gets more accurate evaluation results in comparison with others. This model can be used in Web service management system, Web service selection engine and Web services composition engine.